Running head: FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT
On social network sites (SNS), people are increasingly confronted with news content—even if
they have not actively been looking for it. Although it is widely recognized that SNS have
become a main driver for such incidental news exposure, we know little about the factors that
influence whether users engage with news encountered on SNS. Thus, this study investigates
under which conditions incidental news exposure becomes actual engagement with news by
asking how both the perception of the news post and general news usage patterns influence the
intention to read news articles encountered on SNS as well as the intention to look for further
information about the covered issues. Building on a mobile forced experience sampling study
consisting of 840 Facebook news encounters reported from 124 participants, we find that news
engagement is mostly determined by participants’ perceived interestingness of and prior
knowledge about the issue of the news post and to a much lesser degree by social factors unique
to SNS (i.e., feelings towards the spreader of the news). In contrast, no influence of content-
independent news usage patterns on news engagement could be observed.
Keywords: social network sites, Facebook, online news, incidental news exposure, news
engagement, mobile experience sampling method (MESM)
This is a PDF file of an unedited manuscript that has been accepted for publication in
Computers in Human Behavior: https://doi.org/10.1016/j.chb.2017.06.041
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 1
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
Over the past few years, the ways in which people keep up with what is going on in the
world around them have changed rapidly. Especially social network sites (SNS) like Facebook
have affected how individuals find, consume, and engage with news content. In today’s media
environment, it happens more and more frequently that SNS users encounter news content during
their SNS usage even though they have not been actively looking for it. Imagine a Facebook user
who just wants to post some pictures of her weekend trip, but—while browsing through her news
feed—sees that her best friend liked an article of The New York Times, notices her university
shared a link to a TIME report or stumbles over new content of MSNBC in her timeline, whose
fan page she had liked a few months ago. Hence, the boundaries between (political) news and
mundane everyday life experiences, that is, public and private sphere, are getting more and more
porous (Brundidge, 2010). Although it is widely recognized that SNS have become a main driver
for such incidental news exposure (Gottfried & Shearer, 2016; Mitchell, Gottfried, Barthel, &
Shearer, 2016; Newman, Fletcher, Levy, & Nielsen, 2016), we still know little about the factors
that influence whether users actually engage with news encountered on SNS—or decide to
ignore it. While incidental news exposure neither is a new phenomenon nor exclusively found
online, SNS have dramatically changed the frequency and magnitude of this phenomenon.
According to Pew Research Center polls, 62% of Facebook news users get news on the site
mostly by chance—and not because they were actively trying to catch up with recent
developments (Gottfried & Shearer, 2016). At the same time, SNS, being social media,
systematically add a layer of social cues to incidental news exposure. Hence, additionally to first
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 2
impressions about the news article itself, users’ feelings towards the person, group, or
organization that recommended/spread the article might alter their willingness to engage with
news encountered on SNS.
Given the increasing relevance of incidental news exposure on SNS, our study seeks to
understand under which conditions incidental news exposure on SNS becomes actual
engagement with news. We do so by building on the curated flows framework (Thorson &
Wells, 2016) and asking how both the (content-dependent) perception of the news post and
(content-independent) general news usage patterns influence the intention to read news articles
encountered on Facebook as well as the intention to look for further information about the
covered issues. By doing so, we aim to fill a gap in the literature that—albeit having already
addressed the effects of incidental news exposure on SNS, particularly the relationship with
political participation—did not dig deeper into the phenomenon itself, that is, the intermediate
step from (mere) exposure to engagement with news that users incidentally encounter on SNS.
Empirically, we build on a mobile forced experience sampling study consisting of 840 Facebook
news encounters reported from 124 German participants in October 2016.
2. From Incidental Exposure to Engagement with News
During the past decade, SNS have become a constitutive part of online news distribution
and consumption (Mitchell & Page, 2015). As early as 2012, all US newspapers with a weekday
circulation of more than 100,000 were using SNS as a means to deliver their content (Ju, Jeong,
& Chyi, 2014)—nowadays, five years later, a news provider without a Facebook or Twitter page
seems even more inconceivable. Due to convenient and easy-to-use tools like the “Share” or
“Like” buttons, both online news sites and SNS facilitate and simplify the recommendation and
dissemination of news content among users, which has an effect both on users actively engaging
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 3
in such sharing activities as well as on users exposed to shared news content. Empirical evidence
so far suggests that incidental news exposure on SNS has the potential to foster both political
participation and the learning of political information (Beam, Hutchens, & Hmielowski, 2016;
Bode, 2016; Y. Kim, Chen, & Gil de Zúñiga, 2013; Valeriani & Vaccari, 2015). However, in
order to achieve such beneficial effects, users have to actually engage with the content they
encounter on SNS. In the following sections, we discuss the phenomenon of incidental news
exposure and shed light on factors that might influence the shift from mere exposure to
engagement with news in the context of SNS. As Facebook is “by far the most important
network for finding, reading/watching, and sharing news” (Newman et al., 2016, p. 8; see also
Gottfried & Shearer, 2016, p. 4) and, as a result, the SNS of choice for our empirical study, the
subsequent chapters are generally focused on Facebook. However, most of the discussed
characteristics also apply to similar services like Twitter or local SNS counterparts.
2.1 Incidental News Exposure on the Internet
Incidental exposure to (political) information and news is undoubtedly not a new
phenomenon and thus has been discussed long before the advent of the Internet. Downs (1957),
for example, described two mechanisms of information acquisition that can foster the learning
about public affairs. Focusing on the concept of information cost, he differentiates between
actively collected “sought-for data” and passively acquired “accidental data” (Downs, 1957, p.
223). Defining the latter, he already addresses three essential features that continue to be
distinctive characteristics of incidental forms of information acquisition:
Accidental data are by-products of the non-political activities of a citizen; they accrue to
him without any special effort on his part to find them. Thus, their cost in time is
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 4
ordinarily much lower than that of sought-for data. (Downs, 1957, p. 223, emphasis by
In the ‘offline world’ people could, for example, obtain such accidental data about public
affairs while passing a news kiosk and glancing at a headline, while talking with a neighbor1, or
while waiting for their favorite TV show and inadvertently watching the last five minutes of a
newscast. Research prior to the Internet mainly focused on the potential of television to
encourage incidental or passive learning—commonly with a focus on those not interested in
politics (Blumler, 1970; Blumler & McQuail, 1969; Krugman & Hartley, 1970; Zukin & Snyder,
1984). Following Blumler (1970), television was believed to be particularly influential in
conveying political information in passing due to the sequential presentation of information, the
general appeal of audiovisual stimuli, and the predominantly recreational use of the medium.
Taken together, television made it harder to avoid (political) information—particularly when
compared with newspapers that present different pieces of information simultaneously instead of
chronologically and lack the ‘spell’ of moving pictures. Schönbach and Weaver (1985, p. 173)
termed this the “trap” effect, specifically referring to television’s ability to capture viewers that
are indifferent about politics. In addition, up until the early 1980s, people could only choose
from a handful of networks, making it even harder to escape news programs altogether. In his
seminal work, Post-broadcast Democracy, Prior (2007) argues that with the shift from low-
choice broadcast television to the current high-choice media environment incidental learning of
news, and especially of political information, has become less common for people with no
interest in politics. While cable news and Internet provide “a feast for news junkies” (Prior,
2007, p. 49), people with a preference for entertainment have better options to avoid the news,
which is assumed to lead to more inequalities in terms of political involvement.
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 5
However, while drastically increasing information choices, the Internet also offers an
abundance of possibilities to stumble upon news content by accident, thus structurally reducing
the costs (e.g., in time, in effort) of news acquisition (Brundidge, 2010). Tewksbury and
colleagues (2001) were one of the first to study such incidental news exposure on the World
Wide Web, using Pew Research Center survey data from 1996 and 1998. Their results show that
approximately half of all respondents reported incidental news exposure and that this exposure
was associated with a greater knowledge of current affairs—albeit effect sizes were relatively
small. While the study offers first insights into the phenomenon of incidental news exposure
online, it is limited by the fact that the measurement of incidental news exposure relied on a
simple binary distinction (yes/no). Furthermore, due to reliance on secondary data, the study did
not distinguish between different applications (e.g., search engines, portal websites), thus
providing a rather general look into the phenomenon. More recent studies, however, have
addressed these issues by asking about the frequency of incidental news exposure and examining
specific applications instead of ‘the Internet’ as a whole (Y. Kim et al., 2013; Valeriani &
Vaccari, 2015). Although focusing on political participation as the dependent variable, the results
are quite similar to those of Tewksbury and colleagues (2001): Incidental news exposure is
positively associated with both offline and online political participation. However, while Kim
and colleagues (2013) found that the positive effect of incidental news exposure on online
political participation was weaker for respondents with a preference for entertainment over news,
Valeriani and Vaccari (2015) found that the correlation is stronger among the politically less
interested respondents. Hence, the conclusions range between a rather pessimistic view of
incidental news exposure increasing existing gaps in political participation (Y. Kim et al., 2013)
and a rather optimistic view of incidental news exposure’s ability to reduce said gaps (Valeriani
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 6
& Vaccari, 2015). Considering these differences, one should also note that Kim and colleagues
condensed incidental news exposure via eight different Internet applications into a single index
while Valeriani and Vaccari just focused on incidental news exposure on SNS. However, given
the contradictory empirical evidence so far, it seems beneficial to—before discussing the
potential effects of incidental news exposure on SNS—first unpack and analyze the specific
characteristics of the SNS information environment that might alter the processing of
incidentally encountered news.
2.2 The SNS Information Environment: Rethinking Incidental News Exposure
First, being inadvertently exposed to news on SNS differs from other forms of incidental
news exposure as a) users can personally influence the likelihood of encountering news content
and b) news are increasingly merged with personal social context information. Theorizing on
information exposure in the age of social media, the curated flows framework (Thorson & Wells,
2016) defines five sets of curating actors that shape recent information environments and helps to
understand both the factors influencing the likelihood of incidentally encountering news on SNS
as well as the factors influencing the shift from exposure to engagement. Thorson and Wells
(2016) argue that curation processes in the age of social media are not only undertaken by
journalistic actors, but also by algorithmic filters, strategic communicators, social contacts, and
individual media users. Curation processes of the latter, termed personal curation, are especially
important when considering why some people encounter more news than others. Through
intentional customization of one’s SNS information environment—for example through
following news providers or through prioritizing their posts—users can actively shape what they
would like to see in their news feed. Although the actual content flow is not under the user’s
control (i.e., which articles are linked by a news provider), the likelihood of encountering certain
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 7
types of content can be increased. Thus, while the exposure to a certain article in one’s news feed
is in fact incidental, the act of following a news provider is not. However, even SNS users that
refrain from such an intentional customization are likely to encounter news content through
social curation, that is, curation performed by the human social network to which a given SNS
user is connected. In that case, news content does not flow directly from a news provider to the
user but gets ‘filtered’ through personal contacts (= friends who share or like news articles on
SNS) that might moderate her willingness to engage with a given piece of news. Adding
algorithmic curation to the picture, it can be assumed that posts (or actions) by friends or pages
with whom the user regularly interacts are especially likely to be displayed in her news feed.
Second—albeit not unique for SNS—, being inadvertently exposed to news on SNS does
not necessarily mean being exposed to the whole piece. As the most common practice of news
providers on SNS is to share hyperlinks referring to articles on their websites (Himelboim &
McCreery, 2012; Ju et al., 2014), users generally only see a short illustrated teaser linking to the
original article. While for some forms of factual public affairs knowledge this small piece of
information might just be enough (e.g., “Donald Trump sworn in as 45th president of the United
States”), we would argue that the processing and, ultimately, acquiring of more complex
information—especially of structural (political) knowledge (Eveland & Hutchens, 2008)—
requires more than a short teaser. Indeed, a recent study by Lee and Kim (2017) shows that the
relationship between incidental exposure to news and the recall of news events is fully mediated
by actual exposure to the linked article.
Hence, in order to study socially desirable outcomes of incidental news exposure like
political knowledge or political participation, it seems useful to focus first on the question of
further engagement: What prompts people to click on an incidentally encountered link to a news
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 8
article and engage with its content? Interestingly enough the potential behaviors triggered by the
first encounter with a given news article (like actually reading it or even looking for further
information about the covered issues) are rarely discussed in the literature on incidental news
exposure—although these behaviors will necessarily mediate the relationship with target values
like political knowledge or participation. A gap in the literature that is not only found in the
context of SNS but also in earlier studies focusing on portal websites or search engines that share
the characteristic of exposure to incomplete information on the encounter level. Valuable insight
regarding factors influencing such further engagement can be found in recent studies on selective
exposure in social media environments (see, for example, Beam, 2014; Messing & Westwood,
2014; Winter, Metzger, & Flanagin, 2016). However, these studies seem only partially
transferable to the study of incidental news exposure on SNS. For one thing, these studies have
focused on metric social cues (“3.000 people like that”) instead of personal social cues (“[Your
Friend] shared TIME’s post”), thus dismissing one of SNS’s distinct characteristics. Aside from
that, but closely related, they relied on predefined experimental stimuli instead of users’ actual
SNS experience, thus hampering external validity. Taken together, the unique characteristics of
the SNS (information) environment—individual customization/experience and personal social
context information—have not been adequately addressed yet.
2.3 Factors Influencing the Shift from Exposure to Engagement
Following the curated flows framework (Thorson & Wells, 2016), the shift from SNS news
exposure to further engagement can both be influenced by factors tied to a specific content (e.g.,
in the case of social curation) and factors independent of a specific piece of content (e.g., in the
case of personal curation). In the following, we will discuss the potential of such content-
dependent and content-independent factors in guiding the intention to engage with news posts
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 9
encountered on SNS. We do not claim that these factors cover every relevant aspect in the
process of incidental news exposure. Rather, in this initial study, we focus on more overriding
factors—especially in the area of content-dependent factors—while excluding more fine-grained
perceptions (e.g., of visual stimuli or content popularity). Thus, the discussion should be seen as
a starting point for addressing and indexing additional influencing factors.
Content-dependent Factors. In the context of incidental news exposure on SNS, we
define content-dependent factors as all perceptions that are directly related to a specific news
post a user encounters in her news feed. As these perceptions are subjective in nature, a news
post does not have objective, inherent qualities, but is interpreted by the user according to her
momentary emotional condition, general beliefs, and predispositions. Psychological studies on
human memory and learning, for example, have shown that prior knowledge affects selective
attention in that knowledge-relevant features are being fixated more often than irrelevant ones (S.
Kim & Rehder, 2011). However, applying this to incidental news exposure on SNS, it is an
empirical question whether this initial attention also sparks further engagement. On the one hand,
prior knowledge might inhibit both the intention to read the article that is being referred to in the
news post as well as the intention to look for further information, because the user feels satiated
and does not expect to find any new information. On the other hand, prior knowledge might
facilitate this intention if a user seeks to verify or reconsider the adequacy of her information or
wants to acquire even more knowledge on the topic. Like prior knowledge, users’ topical interest
is another content-dependent factor that should affect both initial attention to and further
engagement with news posts. According to Deci and Ryan (1985, p. 34) interest “plays an
important directive role in intrinsically motivated behavior in that people naturally approach
activities that interest them”. Hence, if a user is highly interested in the topic that the news post
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 10
suggests the linked article to be about, she should be more inclined to engage with it. In fact, a
recent study by Mummolo (2016) shows that topical interest drastically increased the probability
of selecting news items—even when sources were politically dissonant. Building on recent
selective exposure research, the list of content-dependent factors influencing further engagement
with incidentally encountered news posts could be further extended by including concepts like
perceived credibility (e.g., Winter & Krämer, 2014), informational utility (e.g., Knobloch-
Westerwick & Kleinman, 2012), or attitude-consistency (e.g., Winter et al., 2016).
While all the factors mentioned above also apply to engagement decisions in more
traditional media, another content-dependent factor—the perception of the information
intermediary, that is, the news spreader associated with the post—adds a social dimension to
news perceptions on SNS. Focusing on incidental news exposure triggered by posts or
(inter)actions of friends (i.e., social curation, see Thorson & Wells, 2016), the user’s perception
of the friend that visibly shares or likes a news post might alter her willingness to engage with it.
Recent research supports this idea: Kwon, Stefanone, and Barnett (2014), for example, showed
that social contagion induced from SNS exposure was able to increase favorability assessments
for specific ideas. Similarly, Turcotte and colleagues (2015) found that SNS recommendations
from friends perceived as opinion leaders led to more trust in the recommended media outlet and
increased the user’s likelihood of seeking news from that outlet in the future, while Anspach
(2017) uncovered that online endorsements by Facebook friends not only serve as relevant
factors when deciding which content to consume, but are also able to outweigh partisan
selectivity. More generally, it seems reasonable to assume that a user will be more inclined to
engage with a (news) post by a friend she has positive feelings towards compared to a
recommendation by a SNS acquaintance she does not cherish equally. Along these lines,
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 11
Messing and Westwood (2013) found that people were 50% more likely to attend to media
recommended by a strong tie compared to media recommended by a weak tie. Likewise, we
assume that the general feelings people have when seeing posts from certain SNS contacts can be
a relevant factor when it comes to engagement decisions. Of course, the occurrence of social
curation is highly connected with algorithmic curation (Thorson & Wells, 2016): As the
Facebook algorithm prioritizes (news) content that is shared by people from the user’s network
(“Friends and family come first”, see Mosseri, 2016), news shared by (close) friends might not
only be perceived as more relevant but also be encountered more often in the first place. Both
mechanisms are likely to lead to lower exposure to ideological diverse content (Bakshy,
Messing, & Adamic, 2015).
Content-independent Factors. As opposed to content-dependent factors, we define
content-independent factors as general patterns or user characteristics that are not directly related
to the encountered news post. In the context of news use, it seems particularly likely that general
news usage patterns might affect how users deal with news on SNS. It can be assumed, for
example, that users that frequently engage with news outside of SNS are also more likely to
engage with news posts encountered there due to a general preference for current affairs content.
Likewise, users engaging in personal curation, that is, intentional customization of their SNS
news environment (e.g., through following news providers, see Thorson & Wells, 2016), might
be more inclined to click on links to news articles. Supporting these assumptions, a recent survey
from the Pew Research Center (Mitchell et al., 2016) shows that users who are actively seeking
news out—compared with those for whom news are only an incidental part of their online
experience—are more interested in news overall. This suggests that their engagement decisions
might be less selective in terms of covered issues or topics, thus prompting higher levels of
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 12
overall news engagement. Again, many comparable content-independent factors that have been
studied or utilized within the wider domain of selective exposure research (e.g., need for
orientation [Camaj, 2014] or need for cognition [Winter & Krämer, 2012]) could eventually be
considered when investigating the shift from exposure to engagement.
As a first step, derived from the characteristics of the SNS information environment and
existing empirical research, this study seeks to investigate an initial set of factors that might
affect whether users decide to engage with incidentally encountered news posts. We therefore
(I) How do the content-dependent perception of the news post (feelings towards the
spreader; prior knowledge; topical interest) and users’ content-independent general
news usage patterns (frequency of news use; personal curation of news on Facebook)
influence the intention to read news articles encountered on Facebook?
(II) How do the content-dependent perception of the news post (feelings towards the
spreader; prior knowledge; topical interest) and users’ content-independent general
news usage patterns (frequency of news use; personal curation of news on Facebook)
influence the intention to search for further information about the covered issues?
3.1 Design and Sample
Data for our investigation were collected in a naturalistic setting via a mobile forced
experience sampling study accompanied by an online pre-questionnaire among German-speaking
Facebook users in October 2016. This approach is based on the Experience Sampling Method
developed in sociological research on the everyday life (Hektner, Schmidt, & Csikszentmihalyi,
2007). Experience sampling gives in situ access to everyday usage situations of (mobile) media,
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 13
thus allowing to collect data within or close to the context that is studied (Abdesslem, Parris, &
Henderson, 2010; Karnowski, 2013). We adapted this approach in two ways: (1) Alike several
other recent studies (e.g., Bayer, Ellison, Schoenebeck, & Falk, 2016; Struckmann & Karnowski,
2016)—but in contrast to the original Experience Sampling Method—data was gathered using
participants’ personal smartphones instead of paper-and-pencil questionnaires and beepers, and
(2) we did not collect in situ data on a natural everyday usage situation, but induced a usage
situation in the sense of a “forced experience” (analogous to forced exposure studies).
Consequently, we are terming our approach a mobile forced experience sampling study. This
approach allowed us to—although forcing participants into the experience—base our study on a
sample of naturalistic news encounters.
Participants were recruited using a non-commercial online access panel (SoSci Panel).
Although lacking representativeness, the SoSci Panel provides more heterogeneity than regular
student convenience samples, especially regarding education, age, and gender (Leiner, 2016).
Participation in the study was voluntary, unpaid, and participants were guaranteed complete
confidentiality regarding the obtained data. As an incentive to participate, all participants were
furthermore included in a lottery of four vouchers, worth 50 euros each, for an online retail store.
The different stages of the multi-level survey process are illustrated in Figure 1. First,
invitations to take part in the study were sent to 4,250 panelists, of which 4,194 could be
delivered. 788 participants followed the invitation, of which 306 agreed to want to take part in
the study (first opt-in through providing cell phone number and e-mail address), indicated to be
able to take part in the study (preconditions: mobile Facebook use; smartphone with mobile
Internet connection), and completed the pre-questionnaire. Data obtained by the pre-
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 14
questionnaire included information about participants’ general (content-independent) news usage
patterns as well as sociodemographic characteristics (see below for measures).
Figure 1. Stages of the survey process.
Besides, a short instructional video (1:30 min) was integrated into the pre-questionnaire.
The video demonstrated the procedure of responding to the daily questionnaire, thus showing
participants what to expect from the study.2 Second, the participants were asked in an e-mail to
confirm their first opt-in decision (second opt-in). This confirmed opt-in was necessary to ensure
a) that the participants provided a valid e-mail address, and b), even more importantly, that a
third party was not able to subscribe someone else—accidentally or intentionally—to the study.
238 participants confirmed their first opt-in and were then asked to choose a two-hour time
• 4,250 participants of the SoSci Panel were invited to take part in the
study; 4,194 of the invitations could be delivered
• 788 participants followed the invitation, of which 306 agreed to take part
in the study (first opt-in) and completed the pre-questionnaire
• 238 participants confirmed their opt-in decision after receiving an e-mail
(second opt-in); they were then asked to choose the time frame for the
daily text message dispatch
•The remaining participants received a reminder e-mail that was sent one
day before the daily text message dispatch started
•The remaining participants received a daily text message over a period of
eight days that contained an instruction and the link to the questionnaire
•124 participants completed the daily questionnaire at least three times
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 15
frame between 08.00 a.m. and 10.00 p.m. for the daily text message dispatch. We decided to let
the participants choose the time frame themselves to assure that they were in a situation that
allowed to respond in a timely manner. Third, one day before the daily text message dispatch
started, a reminder e-mail was sent to the participants that included a thank-you note as well as,
once more, a link to the instructional video mentioned above. After that, participants received a
daily text message over a period of eight days that contained a short instruction and the link to
the questionnaire that was accessible via participants’ mobile internet connection. The instruction
was written as follows:
Hello, thank you for participating in our study. Please open Facebook and scroll down to the first news post
in your news feed. After that, please open the following questionnaire: [Link].
Hence, as explained in the video (and always accessible as a detailed written instruction
in the daily questionnaire, too) participants were asked to open Facebook on their smartphone
and look for the first post that contained news.3 For this study, we opted not to define news posts
(a problematic task per se, see Vraga, Bode, Smithson, & Troller-Renfree, 2016), but rather went
with participants’ perceptions. Consequently, participants decided for themselves which posts
they considered to be news posts. However, we provided examples of topic areas such as
politics, sports, celebrity news, and the economy, emphasizing that both hard news and soft news
could be defined as news.
Participants were asked to take a close look at the post and remember its topic
respectively the covered issues. Afterwards, they completed the short daily questionnaire. Data
obtained by the daily questionnaire included a short description of the encountered post (as
provided by the participants), variables addressing the content-dependent perception of the news
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 16
post (feelings towards the spreader; prior knowledge; topical interest), and the central dependent
variables (intention to read the news article, intention to look for further information).
As stated above, our initial sample consisted of 238 participants that confirmed the
second opt-in. Table 1 shows how many participants completed the daily questionnaire on how
many of the eight days. After filtering out all participants that did not finish the entire daily
questionnaire on at least three of the eight days, data of 124 participants and 840 completed daily
questionnaires could be used for analysis.
Table 1. Response Behavior (Daily Questionnaire)
Number of days (d)
on which the questionnaire
0 1 2 3 4 5 6 7 8
Number of participants
that completed the
questionnaire on d day(s)
74 32 8 3 10 10 17 33 51
Demographic characteristics of the sample included age (M = 32.47, SD = 10.49), gender
(72.7% female) and educational level (55.6% with a university degree, 32.3% with a higher
education entrance qualification and only 12.1% with lower educational qualifications).
Participants were mainly heavy users of Facebook, with 84% accessing the site daily. A similar
picture could be observed for participants’ news use: 86% use news on a daily basis.
Measures in the pre-questionnaire. Measures in the pre-questionnaire were focused on
participants’ general news usage patterns, thus covering the content-independent factors that
might affect whether users decide to engage with incidentally encountered news posts or not.
Based on Hölig and Hasebrink (2016), frequency of news use was assessed by asking participants
how often they usually read, hear, or watch current news using a 10-point scale ranging from
“never” to “more frequently than hourly” (M = 6.39, SD = 1.21). Following theoretical
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 17
considerations by Thorson and Wells (2016), participants’ amount of personal curation of news
on Facebook was assessed by asking them about how many news providers they are following
on Facebook with the five possible answers being 1=none, 2=up to five, 3=six to ten, 4=eleven to
15 and 5=more than 15 (Mdn = 2).
Measures in the daily questionnaire. Measures in the daily questionnaire were focused
on participants’ subjective perception of the news post, thus covering relevant content-dependent
factors that might affect whether users decide to engage with incidentally encountered news
posts or not. As stated above, the daily questionnaires were administered in the form of an
adapted experience sampling study. Consequently, the length of the daily questionnaires is
limited to very few questions (see Karnowski, 2013; Kubey, Larson, & Csikszentmihalyi, 1996),
making it necessary to mostly rely on single-item measures instead of generally more reliable
Because we assumed that the general feelings people have when seeing posts from
certain Facebook friends—or from groups or (media) organizations they are following—
influence their news engagement, we asked participants what kind of feelings they have when
they see a post from the respective friend/group/organization on a scale from 1=negative feelings
to 5=positive feelings (feelings towards the spreader). Because the information intermediary
(‘news spreader’) is dependent on the specific situation, this question was worded differently
according to whether the participant had indicated that the news post was shared by a friend,
group, or organization. However, as the general direction of effects should be the same for all
these scenarios, we combined the feelings towards the friend/group/organization to a single
variable (M = 3.41, SD = 1.02). Participants’ topical interest was assessed by asking “How much
interest do you have in the topic of the news post?”, with the scale ranging from 1=no interest at
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 18
all to 5=much interest (M = 3.11, SD = 1.28), while participants’ prior knowledge was measured
by asking “Is the topic of the article new to you or do you already have prior knowledge about
it?” on a scale from 1=The topic is completely new to me to 5=I already have a lot of prior
knowledge about it (M = 2.55, SD = 1.38).
The dependent variables intention to read the news article and intention to look for
further information were assessed with the following questions: “If you think about how you are
using Facebook: How likely is it that you give the news article that is mentioned in the post more
than a glance?” (M = 3.00, SD = 1.44) and “How likely is it that you will look for further
information on the topic?” (M = 2.55, SD = 1.38), both on a scale from 1=very unlikely to
5=very likely. As recent research points to the ambiguity of the behavior ‘clicking on news links’
(Groot Kormelink & Costera Meijer, 2017), we refrained from assessing participants’ actual
behavior and decided to ask more specifically about the likelihood of the engagement behaviors
we were interested in.
3.3 Analytical Strategy
The combination of a (single) pre-questionnaire and (multiple) daily questionnaires per
participant as well as the fact that our independent variables are located both on the level of the
user and the level of the news post, results in a nested data structure. Hence, a multilevel
approach is appropriate to answer our research questions. The first level of our model (i.e., level
of the news post) is represented by the data gathered in the daily questionnaires, which were
filled in at least three times by each participant (level-1 measures: feelings towards the spreader,
topical interest, prior knowledge), while the second level (i.e., level of the user) consists of data
derived from the pre-questionnaire (level-2 measures: frequency of news use, personal curation
of news on Facebook). Our analysis is based on a model with randomly varying intercepts for
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 19
each participant/level-2 group (random intercept model, see Robson & Pevalin, 2015). Given the
limited number of observations on the second level, we confine our analyses to a rather simple
multilevel model that allows for accurate testing of our hypotheses but refrains from modelling
possible further effects.
We used two separate regression models for each of the two dependent variables. The
level 1 models explained either the intention to read the news article or the intention to look for
further information by including the participants’ feelings towards the spreader, her topical
interest, and prior knowledge. Group affiliation (i.e., the user/participant) served as a random
variable in this model. In the next step, the group intercepts were then included as outcome
variables in both level 2 regression equations (similar to a means model, see Raudenbush &
Bryk, 2002). All models were calculated using maximum likelihood estimation, allowing for
comparisons of model fit.
In a first step, we investigated whether the participants differed in their intention to read news
articles they encountered on Facebook and whether they were inclined to look for further
information about the issues discussed. Single-level one-way analyses of variance showed highly
significant differences between participants for both dependent measures, FReading (1, 123) = 2.90,
p < .001; FFurtherInformation (1, 123) = 2.52, p < .001.
As it is common in multilevel modelling, we started from a baseline model containing
only our level-1 measures and controlling for interpersonal differences by including the
participants’ IDs as a grouping variable within which the intercept was allowed to vary randomly
per person. Focusing on the intention to read news articles that were incidentally encountered on
Facebook, the person itself (i.e., the grouping variable ID) accounts for 22% of variance in the
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 20
outcome (R2 = 0.22). As shown in Table 2, all level-1 measures contribute significantly to the
intention to read news articles with topical interest being the most important predictor (b = 0.85,
SE = 0.03, p < .001). Feelings towards the spreader (i.e., the person, group, or [media]
organization that visibly shared or liked the news post) also show a positive impact on reading
intention (b = 0.10, SE = 0.03, p < .01), while prior knowledge concerning the topic of the news
post is negatively associated with reading intention (b = −0.10, SE = 0.03, p < .001). Hence, the
more interested a user is in the topic discussed in an article, the less she already knows about the
topic, and the more positive feelings she has towards the “spreader” of the news, the more likely
she is to engage with a news article encountered on Facebook.
In a second step, we added our level-2 measures, resulting in a small, but significant
increase in model fit (χ2(1)= 77.17, p < .001). However, none of the level-2 measures (frequency
of news use, personal curation of news on Facebook) had a significant effect on the intention to
read the respective news articles (see Table 2). Therefore, our results suggest that the intention to
read news articles one has incidentally encountered on SNS is mainly driven by content-
dependent characteristics such as the perceived interestingness of the article, prior knowledge of
the topic, and—to a lesser degree—the feelings towards the “spreader”.
Regarding the predictors of intention to look for further information, a slightly different
picture emerges. In this case, only perceived interestingness contributes significantly to the
outcome (b = 0.78, SE = 0.03, p < .001), whereas feelings towards the spreader of the article as
well as prior knowledge have neither meaningful nor significant effects. As in the model
reported above, adding the level-2 measures leads to a small, but significant increase in model fit
(χ2(1) = 57.21, p < .001). However, none of the level-2 measures are associated with the
intention to look for further information about the issues covered in a news article one has
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 21
encountered on Facebook. Taken together, general news usage patterns like frequency of news
use and personal curation of news on Facebook neither have an impact on users’ intention to read
news articles encountered on Facebook nor to look for further information about the covered
Table 2. Explaining Intentions to Read News Articles Encountered on Facebook and to Look for
Intention to read news
Intention to look for
Level 1 (Perception of the news post)
Feelings towards spreader
Level 2 (General news usage patterns)
Frequency of news use
Personal curation of news on Facebook
Note. Unstandardized coefficients with standard errors in parentheses; number of observations = 802, number
of groups = 124.
The present study attempted to understand under which conditions incidental news
exposure on social network sites (SNS) results in actual engagement with news by asking how
both the (content-dependent) perception of the news post and (content-independent) general
news usage patterns influence the intention to read news articles encountered on Facebook as
well as the intention to look for further information about the covered issues. It does so in a
unique way by building on data from natural SNS usage situations instead of relying on forced
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 22
exposure experiments or retrospective surveys. To uncover possible antecedents of SNS news
engagement, a mobile forced experience sampling study accompanied by an online pre-
questionnaire was conducted among German-speaking Facebook users over the course of eight
days. The collected data allowed us to analyze 840 actual Facebook news encounters reported
from 124 participants.
Summing up our results, we first found that only users’ content-dependent perception of
the news post predicted reading intentions. These findings go along with previous investigations
of engagement decisions in more traditional media, having shown that users’ topical interest of
the content as well as their prior knowledge about the topic had the strongest effect on reading
intentions (e.g., Bolsen & Leeper, 2013; Eilders, 2006, p. 14; Mummolo, 2016). However, we
found that social curation processes distinctive of SNS—in the form of users’ feelings towards
the news spreader—also had a significant effect on users’ reading intentions. Hence, we could
add further evidence to previous empirical research (e.g., Anspach, 2017; Turcotte et al., 2015)
suggesting that the contextual social cues added by social media indeed influence the effects of
media content provided via SNS.
Second, focusing on the intention to search for further information about the covered
issues, the influence of topical interest largely remained as described above for reading intention,
whereas the aforementioned influence of prior knowledge and feelings towards the news
spreader disappeared. One could argue that the information added through social curation
processes only seems to influence further news engagement as long as it demands little effort
from the user. As soon as efforts increase, the influence seems to fade. This discrepancy roughly
resembles the processes modelled by dual process theories (e.g., ELM; Petty & Cacioppo, 1986):
Social information added by SNS curation processes seems to act as a peripheral cue to engage
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 23
further with the news content. However, as soon as further effort is needed, the influence of this
peripheral cue disappears and only characteristics associated with central processing—like
interest in the topic—prevail. Future research should try to clarify, both conceptually and
empirically, the relationship between the influence of social curation and the effort the user
needs to invest in order to further engage with news.
Third, our findings did not support any influence of (content-independent) general news
usage patterns on neither the intention to read the news article nor on the intention to look for
further information. As we could see in our data, the variance of further news engagement is
mainly explained on the first level of the specific (content-dependent) news encounter, whereas
including the second level of overarching news usage patterns adds no explanatory power. This
finding thus adds further empirical evidence to existing research (e.g., Struckmann & Karnowski,
2016) showing that media use—such as further engagement with news content—indeed seems to
be mainly influenced by characteristics of the varying situational contexts of media use instead
of overarching patterns. Although we expected that users that frequently engage with news
content outside of SNS as well as users that engage in personal curation on SNS, that is,
intentional customization of their Facebook news environment (e.g., through following news
providers, see Thorson & Wells, 2016), might be generally more inclined to engage with
incidentally encountered news articles, our data do not support this assumption. Instead, the shift
from incidental news exposure to news engagement seems to be mainly influenced by users’
momentary, situation-dependent thoughts and perceptions.
Taking these three main findings together, we see that further news engagement with
incidentally encountered news articles on Facebook is mainly influenced by the specific
situational interplay between the user and her perceptions of the content, particularly her existing
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 24
topical interest. As further news engagement is mediating beneficial effects of incidental news
exposure like the learning of political information (cf. Beam et al., 2016; Bode, 2016; Lee &
Kim, 2017) or political participation (cf. Y. Kim et al., 2013; Valeriani & Vaccari, 2015), we
therefore have to state that the potential of these beneficial effects might be higher for users who
were interested in the news topic they incidentally encountered anyway. Thus, our results are
reminiscent of those of Kim and colleagues (2013) who found that the positive effects of
incidental news exposure on political participation were stronger for people with a preference for
news over entertainment content. Nevertheless, on a more optimistic note, there seems to be an
(albeit small) influence of social information added by SNS curation processes, which might lead
users to not only learn more about topics they had been interested in before, but also to learn
about “something new”, that is, topics they had no prior interest in. To shed further light on these
processes, future research should both focus on more diverse forms of news engagement (e.g.,
recommending or bookmarking articles, see Oh, Bellur, & Sundar, 2015) as well as on
differences between specific kinds of social curation processes (e.g., news recommendations
directly addressing the user vs. mere observations of friends’ news-related SNS behavior). While
our data indicates that users’ content-dependent perceptions of the news post are the prime driver
of news engagement decisions, it might nevertheless be useful to study the influence of users’
overarching traits or trait-like characteristics. Specifically, “news-related” (personality)
characteristics like need for cognition (NFC) or a “news-finds-me perception” (Gil de Zúñiga,
Weeks, & Ardèvol-Abreu, 2017) might alter the likelihood of news engagement. While prior
studies on the relationship between NFC and news-/information-related web usage (Tsfati &
Cappella, 2005; e.g., Tuten & Bosnjak, 2001) suggest the assumption that users with a high NFC
are more likely to engage with incidentally encountered news articles on SNS, users who are
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 25
high in news-finds-me perceptions (= the extent to which people believe they can indirectly stay
informed about public affairs), might actually be frequently exposed to news posts on SNS, but
not motivated to engage with the linked article (cf. Gil de Zúñiga et al., 2017). Alongside this
consideration of further content-independent factors, the list of content-dependent factors
influencing further engagement with incidentally encountered news posts could be extended by
including concepts like perceived credibility of the news post, perceived news values, or stylistic
techniques that are supposed to “lure readers into clicking on the headline” (i.e., clickbait,
Kuiken, Schuth, Spitters, & Marx, 2017, p. 4).
Our findings, however, should be interpreted with caution due to inherent limitations.
First, the recruitment of participants and thus the composition of the sample hamper the
generalizability of our results. While the SoSci Panel generally provides quite heterogeneous
samples, the inevitable self-selection led to a sample of participants that are highly educated and
heavy users of both Facebook and news. A replication of the study with a more diverse sample,
especially including more participants with lower levels of formal education and less interest in
news, would help to gain a better understanding of the processes influencing news engagement
on SNS. Particularly, it can be assumed that the influence of social information added by SNS
curation processes might be more pronounced among people with little interest in news as they
strive for more guidance in assessing the significance of certain issues or topics. Second, due to
our sample size and the associated restrictions for multilevel modelling, we only focused on a
small set of content-dependent and content-independent factors that might influence the shift
from incidental news exposure to engagement and did not test for possible interaction effects.
Future studies therefore should aim for a bigger sample size, allowing to consider further factors
on both levels (see above) which might be influential when it comes to users’ intention to engage
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 26
with incidentally encountered news. Another methodological limitation concerns the reliance on
single-item measures. Although single-item measures are generally inferior to multi-item
measures, they also have considerable advantages, particularly when working with mobile-based
experience sampling questionnaires. While multi-item measures would have burdened
participants unnecessarily, single-item measures—due to their elimination of redundancy—can
reduce “the fatigue, frustration, and boredom associated with answering highly similar questions
repeatedly” (Robins, Hendin, & Trzesniewski, 2001, p. 152), which is an issue especially when
answering questions on one’s smartphone and/or on the go. However, future studies with a
different methodological approach should aim to include more refined, multi-item measures.
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 27
Abdesslem, F. B., Parris, I., & Henderson, T. (2010). Mobile experience sampling: Reaching the
parts of Facebook other methods cannot reach. In Proceedings of the Privacy and
Usability Methods Pow-Wow (PUMP) (pp. 1–8). Dundee, UK.
Anspach, N. M. (2017). The new personal influence: How our Facebook friends influence the
news we read. Political Communication. https://doi.org/10.1080/10584609.2017.1316329
Bakshy, E., Messing, S., & Adamic, L. A. (2015). Exposure to ideologically diverse news and
opinion on Facebook. Science, 348(6239), 1130–1132.
Basil, M. D., & Brown, W. J. (1994). Interpersonal communication in news diffusion: A study of
“Magic” Johnson’s announcement. Journalism Quarterly, 71(2), 305–320.
Bayer, J. B., Ellison, N. B., Schoenebeck, S. Y., & Falk, E. B. (2016). Sharing the small
moments: Ephemeral social interaction on Snapchat. Information, Communication &
Society, 19(7), 956–977. https://doi.org/10.1080/1369118X.2015.1084349
Beam, M. A. (2014). Automating the news: How personalized news recommender system design
choices impact news reception. Communication Research, 41(8), 1019–1041.
Beam, M. A., Hutchens, M. J., & Hmielowski, J. D. (2016). Clicking vs. sharing: The
relationship between online news behaviors and political knowledge. Computers in
Human Behavior, 59, 215–220. https://doi.org/10.1016/j.chb.2016.02.013
Blumler, J. G. (1970). The political effects of television. In J. D. Halloran (Ed.), The effects of
television (pp. 68–104). London: Panther Books.
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 28
Blumler, J. G., & McQuail, D. (1969). Television in politics: Its uses and influences. Chicago:
University Of Chicago Press.
Bode, L. (2016). Political news in the news feed: Learning politics from social media. Mass
Communication and Society, 19(1), 24–48.
Bolsen, T., & Leeper, T. J. (2013). Self-interest and attention to news among issue publics.
Political Communication, 30(3), 329–348.
Brundidge, J. (2010). Political discussion and news use in the contemporary public sphere.
Javnost - The Public, 17(2), 63–81. https://doi.org/10.1080/13183222.2010.11009031
Camaj, L. (2014). Need for orientation, selective exposure, and attribute agenda-setting effects.
Mass Communication and Society, 17(5), 689–712.
Deci, E., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior
(1985th ed.). New York: Plenum Press.
Downs, A. (1957). An economic theory of democracy. New York: Harper & Brothers.
Eilders, C. (2006). News factors and news decisions. Theoretical and methodological advances
in Germany. Communications, 31(1), 5–24.
Eveland, W. P., & Hutchens, M. J. (2008). Political knowledge. In W. Donsbach (Ed.), The
International Encyclopedia of Communication. Blackwell Publishing. Retrieved from
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 29
Gil de Zúñiga, H., Weeks, B., & Ardèvol-Abreu, A. (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.
Gottfried, J., & Shearer, J. (2016). News use across social media platforms 2016. Pew Research
Center. Retrieved from http://www.journalism.org/files/2016/05/PJ_2016.05.26_social-
Groot Kormelink, T., & Costera Meijer, I. (2017). What clicks actually mean: Exploring digital
news user practices. Journalism, Advance online publication.
Hektner, J. M., Schmidt, J. A., & Csikszentmihalyi, M. (2007). Experience sampling method:
Measuring the quality of everyday life. Thousand Oaks, Calif: SAGE.
Himelboim, I., & McCreery, S. (2012). New technology, old practices: Examining news websites
from a professional perspective. Convergence, 18(4), 427–444.
Hölig, S., & Hasebrink, U. (2016). Reuters Institute Digital News Survey 2016 – Ergebnisse für
Deutschland (Arbeitspapiere des Hans-Bredow-Instituts No. 38). Hamburg, Deutschland:
Ju, A., Jeong, S. H., & Chyi, H. I. (2014). Will social media save newspapers? Examining the
effectiveness of Facebook and Twitter as news platforms. Journalism Practice, 8(1), 1–
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 30
Karnowski, V. (2013). Befragung in situ: Die Mobile Experience Sampling Method (MESM). In
W. Möhring & D. Schlütz (Eds.), Handbuch standardisierte Erhebungsverfahren in der
Kommunikationswissenschaft (pp. 235–247). Wiesbaden: Springer.
Kim, S., & Rehder, B. (2011). How prior knowledge affects selective attention during category
learning: An eyetracking study. Memory & Cognition, 39(4), 649–665.
Kim, Y., Chen, H.-T., & Gil de Zúñiga, H. (2013). Stumbling upon news on the internet: Effects
of incidental news exposure and relative entertainment use on political engagement.
Computers in Human Behavior, 29, 2607–2614.
Knobloch-Westerwick, S., & Kleinman, S. B. (2012). Preelection selective exposure:
Confirmation bias versus informational utility. Communication Research, 39(2), 170–
Krugman, H. E., & Hartley, E. L. (1970). Passive learning from television. Public Opinion
Quarterly, 34(2), 184–190. https://doi.org/10.1086/267788
Kubey, R., Larson, R., & Csikszentmihalyi, M. (1996). Experience sampling method applications
to communication research questions. Journal of Communication, 46(2), 99–120.
Kuiken, J., Schuth, A., Spitters, M., & Marx, M. (2017). Effective headlines of newspaper
articles in a digital environment. Digital Journalism, 0(0), 1–15.
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 31
Kwon, K. H., Stefanone, M. A., & Barnett, G. A. (2014). Social network influence on online
behavioral choices. Exploring group formation on social network sites. American
Behavioral Scientist, 58(10), 1345–1360. https://doi.org/10.1177/0002764214527092
Lee, J. K., & Kim, E. (2017). Incidental exposure to news: Predictors in the social media setting
and effects on information gain online. Computers in Human Behavior.
Leiner, D. J. (2016). Our research’s breadth lives on convenience samples. A case study of the
online respondent pool “SoSci Panel.” SC|M Studies in Communication | Media, 5(4),
Messing, S., & Westwood, S. J. (2013, October 21). Friends that matter: How social influence
affects selection in social media. Retrieved from
Messing, S., & Westwood, S. J. (2014). Selective exposure in the age of social media:
Endorsements trump partisan source affiliation when selecting news online.
Communication Research, 41(8), 1042–1063. https://doi.org/10.1177/0093650212466406
Mitchell, A., Gottfried, J., Barthel, M., & Shearer, E. (2016). The modern news consumer. News
attitudes and practices in the digital era. Pew Research Center. Retrieved from
Mitchell, A., & Page, D. (2015). The evolving role of news on Twitter and Facebook. Pew
Research Center. Retrieved from http://www.journalism.org/files/2015/07/Twitter-and-
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 32
Mosseri, A. (2016, June 29). Building a better news feed for you. Retrieved June 6, 2017, from
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., Fletcher, R., Levy, D. A. L., & Nielsen, R. K. (2016). Reuters Institute Digital
News Report 2016. Oxford, UK: Reuters Institute for the Study of Journalism.
Oh, J., Bellur, S., & Sundar, S. S. (2015). Clicking, assessing, immersing, and sharing: An
empirical model of user engagement with interactive media. Communication Research.
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In L.
Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 19, pp. 123–205).
New York: Academic Press.
Prior, M. (2007). Post-broadcast democracy: How media choice increases inequality in political
involvement and polarizes elections. New York: Cambridge University Press. Retrieved
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data
analysis methods. Thousand Oaks, CA: SAGE.
Robins, R. W., Hendin, H. M., & Trzesniewski, K. H. (2001). Measuring global self-esteem:
Construct validation of a single-item measure and the Rosenberg Self-Esteem Scale.
Personality and Social Psychology Bulletin, 27(2), 151–161.
Robson, K., & Pevalin, D. (2015). Multilevel modeling in plain language. Thousand Oaks, CA:
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 33
Schönbach, K., & Weaver, D. H. (1985). Finding the unexpected: Cognitive bonding in a
political campaign. In S. Kraus & R. M. Perloff (Eds.), Mass media and political thought:
An information-processing approach (pp. 157–176). Beverly Hills: SAGE.
Struckmann, S., & Karnowski, V. (2016). News consumption in a changing media ecology: An
MESM-study on mobile news. Telematics and Informatics, 33(2), 309–319.
Tewksbury, D., Weaver, A. J., & Maddex, B. D. (2001). Accidentally informed: Incidental news
exposure on the world wide web. Journalism & Mass Communication Quarterly, 78(3),
Thorson, K., & Wells, C. (2016). Curated flows: A framework for mapping media exposure in
the digital age. Communication Theory, 26(3), 309–328.
Tsfati, Y., & Cappella, J. N. (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.
Turcotte, J., York, C., Irving, J., Scholl, R. M., & Pingree, R. J. (2015). News recommendations
from social media opinion leaders: Effects on media trust and information seeking.
Journal of Computer-Mediated Communication, 20(5), 520–535.
Tuten, T. L., & Bosnjak, M. (2001). Understanding differences in web usage: The role of need
for cognition and the five factor model of personality. Social Behavior and Personality:
An International Journal, 29(4), 391–398. https://doi.org/10.2224/sbp.2001.29.4.391
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 34
Valeriani, A., & Vaccari, C. (2015). Accidental exposure to politics on social media as online
participation equalizer in Germany, Italy, and the United Kingdom. New Media &
Society, 1461444815616223. https://doi.org/10.1177/1461444815616223
Vraga, E. K., Bode, L., Smithson, A.-B., & Troller-Renfree, S. (2016). Blurred lines: Defining
social, news, and political posts on Facebook. Journal of Information Technology &
Politics, 13(3), 272–294. https://doi.org/10.1080/19331681.2016.1160265
Winter, S., & Krämer, N. C. (2012). Selecting science information in web 2.0: How source cues,
message sidedness, and need for cognition influence users’ exposure to blog posts.
Journal of Computer-Mediated Communication, 18(1), 80–96.
Winter, S., & Krämer, N. C. (2014). A question of credibility – Effects of source cues and
recommendations on information selection on news sites and blogs. Communications,
39(4), 435–456. https://doi.org/10.1515/commun-2014-0020
Winter, S., Metzger, M. J., & Flanagin, A. J. (2016). Selective use of news cues: A multiple-
motive perspective on information selection in social media environments. Journal of
Communication, 66(4), 669–693. https://doi.org/10.1111/jcom.12241
Zukin, C., & Snyder, R. (1984). Passive learning: When the media environment is the message.
Public Opinion Quarterly, 48(3), 629–638. https://doi.org/10.1086/268864
FROM INCIDENTAL NEWS EXPOSURE TO NEWS ENGAGEMENT 35
1Interpersonal communication as a means of encountering news seems to be especially
prevalent in times of ‘big’ news events. A meta-analysis of news diffusion studies by Basil and
Brown (1994) suggests that the more important the event, the more people are likely to hear
about it through interpersonal communication.
2To enable blind review, the video will be made available only after the review process is
3Because we asked participants about the first post that contained news, this post was not
necessarily the first post in their news feed. Although we—due to wanting to keep the daily
questionnaire as short and straightforward as possible—did not ask participants for the post’s
exact position, it can be assumed that this approach yields more variance in types of posts than
simply asking about the first post and filtering for news posts ex post.
4All questionnaires were administered in German. All items discussed in the method
section represent English translations of the original items.