#NewsConsumptionIn2017
Young People’s News Repertoires and their Political Implications
Author: Timea Rüb
Supervisor: Jasper van de Pol
Graduate School of Communication
Master’s Programme Communication Science
Master’s Thesis
Student ID: 11181451
Abstract
The present study uses a repertoire approach to analyze young people’s (ages 16 to 25) news consumption habits. Further, it examines the associations between these repertoires and political
knowledge and -efficacy. In doing so, this study closes a gap in the news consumption literature,
which does not presently contain up-to-date studies on the news repertoires of young people and
the political implications of these repertoires. Results from an online-survey of German
16-to-25-year-olds reveal five separate news repertoires: (1) Globally oriented traditionalists, (2) News
junkies, (3) Nationally oriented dabblers, (4) News avoiders, and (5) Online news users. No
significant age-differences are found between the five repertoires. Multiple one-way ANCOVAs
indicate that the repertoires differ significantly in terms of both political knowledge and
-efficacy. Overall, results show that the repertoires which predominantly feature traditional media
are most strongly associated with political knowledge and internal political efficacy. External
efficacy appears to be promoted by both traditional and new-media-centered news repertoires.
Implications for future research on young people’s news repertoires, the nature of their individual components, their democratic implications, as well as their potential predictors are
discussed.
Keywords: News repertoires, young people, news consumption, political knowledge,
Introduction
Young people today are becoming increasingly disengaged from the news (Mindich, 2005).
They follow the news significantly less often than their elders do (Mitchell, Gottfried, Barthel, &
Shearer, 2016), especially by means of traditional media (Huang, 2009; Patterson, 2007;
Vandebosch, Dhoest, & Van den Bulck, 2009). When young people do consume news, they
seem to increasingly consult online sources like social media (Casero-Ripollés, 2012; Pew,
2015). Whether this is a good or bad thing remains to be seen, but it is self-evident that the news
consumption habits of young people have crucial democratic implications. Today’s youth is tomorrow’s future, and the amount and depth of political information they encounter has a profound impact on their abilities to participate meaningfully in the democratic process (de
Vreese & Boomgaarden, 2006; Hao, Wen, & George, 2014; Prior, 2005).
An abundance of scholarship has focused on the news consumption patterns of today’s
youth and their democratic implications (e.g., Bakker & de Vreese, 2011; Baumgartner &
Morris, 2010; Casero-Ripollés, 2012; Hao et al., 2014; Loader, Vromen, & Xenos, 2014;
Poindexter, 2012). However, a large majority of these studies are either single-media-use studies
or studies investigating multiple-media-use without taking a closer look at the relationships
between the media sources used and how these are combined to form distinct
media-use-typologies.
But much can be said for investigating media use in terms of aggregate patterns of
consumption. Media users inevitably combine their media-selections into comprehensive
patterns of overall exposure (Hasebrink & Popp, 2006). Additionally, the introduction of mobile
media technologies and the associated cross-platform content means that modern-day media use
tech-savvy younger generation.
Hasebrink and Popp label such aggregate patterns of media exposure “media repertoires”
(2006, p.369). Specifically, they define the media repertoire as a “comprehensive pattern of
[media] exposure” (p. 369). The repertoire approach to studying media use allows the researcher to ‘build a map’ of the media mix that people use and to identify specific media and content-types which are frequently used in conjunction with each other.
The repertoire approach has been used to identify sets of TV-channels which viewers
regularly watch in combination with each other (Heeter, 1985) and to determine general
media-combination typologies in a cross-media environment (Hasebrink, Jensen, van den Bulck, Hölig,
& Maeseele, 2015; Kim, 2014; Taneja, Webster, Malthouse, & Ksiazek, 2012; van Rees & van
Eijck, 2003). In the context of news media consumption, it has been employed to determine
specific combinations of sources people use to get news, i.e. to determine people’s news
repertoires (Edgerly, 2015; Hasebrink & Schmidt, 2012; Ksiazek, Malthouse, & Webster, 2010;
Lai & Tang, 2015; Lee & Yang, 2014; Schrøder & Kobbernagel, 2010; Swart, Peters, &
Broersma, 2016; Trilling & Schoenbach, 2013; Wolf & Schnauber, 2015; Wolfsfeld, Yarchi, &
Samuel-Azran, 2016; Yuan, 2011).
Unfortunately, very little is known about the news repertoires of young people. To the
author’s best knowledge, only two existing studies investigate the news repertoires of young people (Diddi & LaRose, 2006; Van Cauwenberge, d'Haenens, & Beentjes, 2011). Both can be
considered outdated, especially in light of the rapidly changing modern-day news-media
environment. Furthermore, the first of these studies (Diddi & LaRose, 2006) uses a sample of
American college students in which the teenage age-range is not represented. Similarly, the
uses an exceedingly broad definition of ‘young’ people, which ranges from ages 15 to 34.
Furthermore, neither of the two studies investigates the democratic implications of the identified
news repertoires.
Among these democratic implications are political knowledge and -efficacy, which have
been shown to play a decisive role in explaining a range of predicted democratic participation
behaviors in young citizens (e.g., Amnå, Munck, & Zetterberg, 2004; Carpini & Keeter, 1997).
Moreover, research has shown that different political information-types and -channels
encountered by young people are associated with the experience of different levels of
“confidence in … [their] own political knowledge and its sufficiency to engage the political process…” (Kaid, McKinney, & Tedesco, 2007, p. 1096). This association presumably also holds true for the different types of news repertoires young people subscribe to and their
respective levels of political knowledge and -efficacy.
It therefore appears relevant, both scientifically and societally, to investigate the news
repertoires of young people between the ages of 16 and 25 as well as the democratic implications
of these. Consequently, the following principal research questions guide the present thesis: How
are young people’s (ages 16 to 25) news repertoires structured and to what extent are these repertoires associated with different levels of political knowledge and political efficacy?
Theoretical Framework The Media Repertoire Approach
The repertoire approach to (news) media use moves away from traditional,
displacement-theory-based conceptions of media use, which regard consumption in an expanding media
environment as a zero-sum game, in which time spent on one medium reduces time spent on
media complementarity, which suggests that media outlets should not be regarded as being in
competition with each another, but rather as complementary components within the media
consumption process (Dutta-Bergman, 2004). Media users will select and combine these
complementary media based on their worthwhileness (Schrøder, 2015), i.e. the extent to which
they are worthwhile in their contribution to, and their ‘fit’ within a user’s selection of regularly-consumed media. Accordingly, a media repertoire can be defined as “the collection of media
sources that people regularly use” (Kim, 2014, p. 354) or as a “subset” of media outlets, selected from the extensive array of available choices (Taneja et al., 2012, p. 953).
The inclusive and comprehensive character of the repertoire approach to studying media
use is especially relevant in today’s ever-expanding environment of media convergence and abundant media choice, which inevitably requires audiences to construct small and manageable
subsets of preferred media so that they can make sense of the complex media landscape.
Studying people’s media repertoires. In media repertoire studies, the media variable
under study is not a single variable or the mere sum of different media used. Instead, the media
variables under study combine into a “meaningfully structured composition of media”
(Hasebrink & Domeyer, 2012, p. 760). The components of media repertoires (i.e. their ‘building
blocks’) can have a variety of ‘levels’: A repertoire can be made up of certain media/modalities (e.g., TV, newspapers, Internet, etc.), specific content-types/channels (e.g., infotainment news,
reality TV, etc.), as well as specific media brands (e.g., legacy brands like the Wall Street
Journal). The repertoire approach is not exclusive to any one of these levels (Hasebrink & Popp,
2006). A media repertoire’s structure is characterized by the number of repertoire-components
(breadth), its diversity, and the degree of complementarity between its individual components
their diversity and complementarity. Furthermore, and in contrast to many existing repertoire
studies, which have neglected to characterize repertoires according to more than one
component-level (Ksiazek et al., 2010; Lee & Yang, 2014; Taneja et al., 2012; Van Cauwenberge et al.,
2011), this study uses both the modality and content-type/channel levels to identify and define
repertoires.
It is advantageous to analyze media repertoires in the context of concrete societal groups
because analyses on the aggregate level may hide significant differences in the media-use
patterns of societal subgroups (Hasebrink & Popp, 2006). Young people are an important
societal subgroup whose media repertoires are worth studying. This is because people’s media repertoires are closely related to their life-stage and the configurations of media repertoires
change as people transition from one life-stage to the next (Kim, 2014; van Rees & van Eijck,
2003). The phase of young adulthood is an important life-stage as it encompasses an emerging
independence from the parents and family, the exploration of new life choices, and the creation
of new, self-determined behavioral patterns (Antunovic, Parsons, & Cooke, 2016). Media habits
are a central part of such behavioral patterns: Young people lay the foundations for their media
repertoires whose configurations will remain relatively stable throughout their years of young
adulthood during this time (Diddi & LaRose, 2006; LaRose, 2010). Consequently, studying
young people’s media repertoires is likely to yield meaningful insights, especially when it comes to their democratically consequential news repertoires.
The Structure of Young People’s News Repertoires
For the purposes of this thesis, news repertoires will be defined as people’s selections of
daily news media across multiple available modalities and content-types/channels (Lee & Yang,
Yang, 2014; Trilling & Schoenbach, 2013). While a considerable shortcoming of these studies is
that they focus largely on the news repertoires of adults,1 together, they provide an indication of
the conceivable structures of young people’s news repertoires in the current media environment. For instance, Hasebrink and Schmidt (2012), in their study of the information repertoires
of German citizens, found that some information repertoires exhibited preferences for specific
modalities (e.g., radio), while others could be characterized as more content/channel-driven (e.g.,
classic journalism). Similarly, studies by Ksiazek et al. (2010) and Yuan (2011) showed that
people rely on a variety of different media and content-types/channels to make sense of current
affairs. Kim (2014), in her study of the general media repertoires of Korean citizens, aptly
summarizes these findings when she states, “[S]ome media repertoires are formed around media
… while others are established around content types” (p. 356). This leads to the formulation of the first hypothesis concerning the expected structure of young people’s news repertoires:
H1: Young people’s news repertoires will be structured around both modality and content/channel, with some repertoires being distinguishable as modality-specific and
others as channel/content-specific, while others yet will entail both a variety of
channels/content and modalities.
Furthermore, existing studies on young people’s general news consumption habits
provide indications about potential age-based differences in the structure of young people’s news
repertoires. Van Eijck and van Rees (2000) showed that the ‘older’ group of young adults
(between ages 20 and 30) tends to have an ‘omnivorous’ news media diet that combines a
multitude of heterogeneous media and content-types. Similarly, Casero-Ripollés (2012) found
that the use of a diverse set of media for news consumption grows as young people mature. On
1 As indicated in the introduction, there are only two existing studies on the news repertoires of young people (Diddi
the flip side, Craft, Ashley, and Maksl (2016) showed that teenagers’ news consumption habits
focus on only very few media outlets because deliberate and thorough news-seeking behavior is
not (yet) regarded as necessary and advantageous in this age group. Indeed, the teenagers
interviewed in the study expressed the impression that keeping up with a wide variety of news
through a multitude of outlets is “something for the future” (pp. 151-152).
Younger youth seem to associate thorough and varied news consumption with adulthood
and their eventual incorporation into the labor market and therefore do not consider it pertinent
to them (Casero-Ripollés, 2012). Older youth, on the other hand, regard comprehensive news
consumption as a form of empowerment, enabling them to make everyday decisions as they
become “increasingly responsible for the trajectories of their own lives…” (Van Cauwenberge, d’Haenens, & Beentjes, 2013, p. 375). Consequently, when it comes to the amount and range of news media consumed by young people, there appears to be an age effect. With this in mind, the
second hypothesis concerning the expected structure of young people’s news repertoires is
formulated:
H2: Young adults (ages 20 to 25) will have more diverse news repertoires (both in terms of modality and content/channels) than adolescents (ages 16 to 19).
The literature points to similar age-based differences when it comes to the expected
prominence of traditional and new media in young people’s news repertoires. Generally, news repertoires oriented toward traditional media have been found to be more representative of older
persons (Hasebrink & Schmidt, 2012; Kim, 2014; Lee & Yang, 2014; Taneja et al., 2012) and
this tendency appears to also apply to young people. For example, Choi (2016) states that
younger youth are more likely to have news repertoires that are rich in Internet-based media
sample of Catalonian youth was more likely to express interest in, and usage of traditional media
than the younger respondents (ages 16 and 17).
This more frequent use of traditional media for news consumption among young adults
than among teens is presumably the result of a cohort/generation effect. While today’s young
adults (currently in their twenties) have grown up with traditional news media, today’s teenagers have grown up in an almost entirely digitalized news media environment (Bovill & Livingstone,
2001). In particular, the rise of individualized and less visible modes of news consumption (often
via mobile devices), has meant that these younger youth, while growing up, have not
experienced the use of traditional media in the home in the same way as their older counterparts
(Edgerly, Thorson, Thorson, Vraga, & Bode, 2017). Based on these insights, the third hypothesis
concerning the expected structure of young people’s news repertoires is formulated:
H3: Young adults’ (ages 20 to 25) news repertoires will feature more traditional media
than adolescents’ (ages 16 to 19) news repertoires.
Political Implications of Young People’s News Repertoires
The various news repertoires which young people subscribe to may have important
impacts on their levels of political knowledge and -efficacy, which, in turn, influence their
abilities to effectively participate in the democratic process (de Vreese & Boomgaarden, 2006;
Hao, Wen, & George, 2014; Prior, 2005).
The political implications of people’s news repertoires and general news-consumption habits have been investigated in a number of studies using adult samples. Papathanassopoulos et
al. (2013) found that people who consume news media across a variety of outlets have higher
levels of political knowledge and political efficacy than those who use only a few media outlets
people who use a diverse set of local news media experience an increase in community
informedness and community efficacy. Furthermore, Wolfsfeld et al. (2016) found that a richer
political information repertoire is linked to high levels of political efficacy and knowledge,
because it exposes its users to a variety of different viewpoints and news items and requires that
they spend time engaging with this variety, effectively giving them a “certain confidence in their
ability to have an impact” (p. 2100). These findings are expected to hold for young people as well. Thus, the following hypotheses concerning the associations between the news repertoires of
young people and their respective levels of political knowledge and -efficacy are formulated:
H4a: More diverse news repertoires (both in terms of modality and content/channels) will be positively associated with political knowledge.
H4b: More diverse news repertoires (both in terms of modality and content/channels) will be positively associated with political efficacy.
The prominence of traditional versus new media within young people’s news repertoires may also differentially impact their levels of political knowledge and -efficacy. Lee and Yang
(2014), in their study of the information repertoires of Korean adults, found that participants of
the ‘traditional news-seekers’ repertoire outperformed all other repertoire groups in terms of political knowledge. This is in line with research confirming that people acquire more socially
and politically important information from traditional media than from emerging new media
(e.g., Dimitrova, Shehata, Strömbäck, & Nord, 2011; Tewksbury, 2003; Tran, 2013; Yang &
Grabe, 2011).
When it comes to young people’s feelings of political efficacy, the expected differential influence of news repertoires which prominently feature traditional media, and those which
traditional news media promotes political efficacy in young people (e.g., Siyoung, 2011), a
number of other studies show that the use of emerging news media has a much more positive
influence on young people’s feelings of political efficacy than the use of traditional media (e.g., Gil de Zúñiga, Jung, & Valenzuela, 2012; Gil de Zúñiga, Puig-I-Abril, & Rojas, 2009; Tedesco,
2007). This is attributed to the personalization and communicativeness of new media
technologies (e.g., social media), which are said to empower young people with new ways of
expressing opinions and retrieving personally relevant information (Bennett, 2008, 2012). With
these findings in mind, expectations regarding the differential influence of traditional and
new-media-governed news repertoires on young people’s political knowledge and -efficacy are
formulated in terms of a hypothesis and a research question:
H5: News repertoires which prominently feature traditional media will be
more positively associated with political knowledge than those which prominently or
only feature new media.
RQ1: To what extent are news repertoires which prominently feature traditional versus those which prominently feature new media differentially associated with political
efficacy?
The hypotheses are summarized in Figure 1 below.
Method Research Design
To answer the research question guiding this thesis, a cross-sectional online survey was
conducted. Data were collected from April 26th to May 14th, 2017, using the research software
Qualtrics. Respondents had to be between the ages of 16 and 25 and be either German or living
in Germany at the time of data collection.2 As done in the majority of existing media-repertoire
studies (e.g., Van Cauwenberge et al., 2011; Yuan, 2011), an online survey design was selected.
Respondents were recruited through the researcher’s extended personal network via Facebook and email, through the German survey-sharing group “Umfragen für Studienarbeiten”
(“Surveys for theses”) on Facebook, and through the website “SuveyCircle.com,” used (mainly) by university students to share surveys. Upon survey-completion, respondents had the
opportunity to enter their email addresses for a chance to win one of three Amazon gift cards
each worth 10€.
Sample
Data were collected from a convenience sample of 158 respondents,3 68.4% of which
were female, 30.4% were male, and 1.3% of which did not indicate their gender. Respondents
ranged from 16 to 25 years of age (M = 22.30, SD = 2.69). The age-distribution in the sample
was skewed toward the 21-to-25 age-bracket (82.3%). Furthermore, most respondents (89.2%)
were German, with the rest reporting a diverse number of nationalities.4 Regarding education
level, 33.5% of respondents indicated that they had completed or were currently pursuing an
2 This ensured that both the adolescent and the young adult age-ranges were covered. Also, the upper cut-off age
was not so high that the oldest respondents could not reasonably be considered ‘young’ people anymore (in contrast to the age-ranges used in the two existing studies on young people’s news repertoires: college-age and ages 15 to 34 (Diddi & LaRose, 2006; Van Cauwenberge et al., 2011)).
3 Of 204 responses initially gathered, 46 were dropped because the respective respondents did not complete the
survey in its entirety, leaving a final analytic sample of 158 respondents.
intermediate school leaving certificate or high school diploma, 8.2% had completed or were
pursuing some form of vocational training, and 58.3% reported completing or pursuing a
Bachelor’s or Master’s degree or equivalent level qualification.
Both females and young adults (20-to-25-year-olds) were overrepresented in the sample:
The 2011 census indicates that 48.9% of German 16-to-25-year-olds are female and that there
are around 1.23% more young adults than adolescents (bpb, 2012). In terms of education, the
sample was slightly more representative of 16-to-25-year-old Germans, with the most recent
statistics indicating that 53% of Germans aged 18 or above are enrolled in or have completed a
university program (AGJ, 2017).
Measures
News media use. In order to later identify their news repertoires, respondents were asked
to indicate the typical frequency with which they used an approximately exhaustive list of 27
information sources. This method has been successfully utilized by a number of existing media
repertoire studies (e.g., Edgerly, 2015; Ksiazek et al., 2010). Specifically, the battery of 27 news
media items used was modelled after the 24-item list used in a 2012 study by Hasebrink and
Schmidt, which examined the news repertoires of the German adult population. The 27 items
used in the present study were formulated at the content/channel-level (e.g., “main news
broadcasts on public service television” instead of “television”) with the goal of covering both news content-types/channels and modalities and thereby later identifying news repertoires that
are as nuanced as possible.
To achieve the most exhaustive list of news media items possible, pilot interviews were
asked about the sources they regularly consulted to attain current affairs information.5 Based on
these interviews, the final battery of 27 information sources was compiled.6 Respondents were
asked to indicate how often they typically used each information source on a 7-point scale (1 =
“never,” 2 = “less than once a month,” 3 = “1 to 3 times per month,” 4 = “1 to 3 times per week,” 5 = “4 to 6 times per week,” 6 = “daily,” 7 = “several times a day”). For the full table of
descriptive statistics of the 27 media use variables, see Appendix A.
Political knowledge. Political knowledge was measured using four ANES-style
political-system and current-affairs knowledge items pertaining to the German political political-system and
German current affairs (e.g., “Which vote determines the strength of the parties in the Bundestag?”).Various studies have demonstrated that these types of questions are reliable
measures of political knowledge (e.g. Carpini & Keeter, 1993, 1997). Each question had four
answer options (two items included a “don’t know” option). Correct answers were scored as “1” and incorrect answers as “0” (“don’t know” answers were given as score of “0”) and a political knowledge scale-score, ranging from 0 (no answers correct) to 4 (all answers correct) was
created by summing across the four items (M = 2.39, SD = 1.08).
Political efficacy. Both internal and external political efficacy were measured in the
survey. Internal political efficacy i.e. “the individual's self-perception that he or she is capable
of understanding politics and competent enough to participate in political acts such as voting” (Miller, Miller, & Schneider, 1980, p. 253) – was measured using the three items with the highest factor loadings from Niemi, Craig and Mattei’s (1991) internal efficacy scale (see
5 Questions were posed using a specific technique by which interviewees were first asked to list their everyday news
sources and then asked open, current-events-related questions designed to help them identify habitual news sources they were not immediately aware of (e.g., “Did you hear about the recent protests against the appearance of Turkish leaders in Germany to campaign for the upcoming referendum in Turkey? If so, where did you see this/hear about this first? Did you then seek out further information about this? If so, where and how?”).
6 Five of the 27 items were television-based, two were radio/podcast-oriented, six were newspaper-based (online and
Appendix B).
External political efficacy i.e. an individual’s feeling of ‘system responsiveness’ or her
belief that the public can influence political outcomes (Craig & Maggiotto, 1982; Miller et al., 1980) was measured using the the three items from Craig and Maggiotto’s (1982) external effiacy scale which had the highest factor loadings and were simultaneously applicable to the present study’s young target group (see Appendix C). Respondents indicated their level of agreement with both the internal and external efficacy items on a 5-point Likert scale (1 =
“strongly disagree,” 5 = “strongly agree”).
Two separate scale-scores were created.7 The internal political efficacy scale-score was
created by averaging across the three internal efficacy items, resulting in an acceptable
Cronbach’s alpha of .78 (M = 3.35, SD = 0.83). For the external political efficacy scale-score, all three external efficacy items were first reverse-scored, and then averaged, so as to achieve a
scale on which high scores indicated high levels of external political efficacy and low scores
indicated the reverse (α = .71, M = 3.18, SD = 0.71).
Political orientation. Political orientation was included in the survey as a control
variable. It was measured using a single item asking respondents to position themselves on a
scale ranging from 0 (“strongly left-leaning”) to 10 (“strongly right-leaning”) (M = 3.75, SD =
1.69).
7 A principle component analysis (PCA) was conducted using the (in total) six internal and external political efficacy
items to determine whether they indeed represented two separate constructs/factors. The results suggested a two-factor solution: Two components had eigenvalues above 1 (2.54 and 1.48) and the scree plot indicated a clear inflection-point after component 2. As was expected, the three internal political efficacy items loaded on one factor and the three external political efficacy items loaded on the other (no cross-loadings appeared).
Data Analysis Plan
To identify the news repertoires of the young people in this study, latent profile analysis
(LPA) was applied. LPA is an exploratory statistical method for the identification of unmeasured
and hidden (latent) groups in a set of data (Oberski, 2016). The latent profiles describe
prototypical subgroups that characterize the response patterns in a number of variables. LPA is a
model-based approach that aims to reveal latent groups based on the means of continuous
observed variables, making it suitable for the present analysis. In this way, it resembles the factor
analytic approach, except that the clusters identified through LPA are categorical constructs
rather than continuous dimensions. LPA is the counterpart to the more well-known method of
latent class analysis, which is used when observed variables are categorical (Hagenaars &
McCutcheon, 2002) (see Appendix D for a thorough explanation of the LPA method).
LPA was applied in the present study to identify the news repertoires (latent profiles) that
explain the relationships between the 27 media-channel-variables measured. The resulting news
repertoires were then interpreted according to their respective conditional response means and
the demographic characteristics of their members. After thorough interpretation of each
repertoire, multiple one-way ANCOVAs were used to examine the associations between
repertoire-membership and political knowledge and -efficacy.
Results Determining News Repertoires (H1)
LPA was conducted in R using the mclust package (Scrucca, Fop, Murphy, & Raftery,
2016). Mclust immediately provides the best-fitting cluster solution using the Bayesian
information criterion (BIC) by default (Fraley & Raftery, 1998; Scrucca et al., 2016). Therefore,
Nevertheless, the five-cluster solution initially provided by mclust was compared to the next best
cluster solutions identified (four and six clusters) using the BIC (Fraley & Raftery, 1998) and the
integrated complete-data likelihood criterion (ICL) (Biernacki, Celeux, & Govaert, 2000).
Both BIC and ICL evidenced that the initial five-cluster solution was the optimal fit for
the data as it yielded the highest values on both fit indices (summarized in Table 1).8
Additionally, the five-cluster solution identified by mclust was replicated in MPlus 8 (Muthén &
Muthén, 2017),9 providing further evidence of the appropriateness of the five-cluster model.
Table 1.
Model Fit Indices
Solution BIC ICL
4 clusters -15660.49 -15673.79
5 clusters -15615.87 -15630.16
6 clusters -15659.15 -15669.66
Note. BIC = Bayesian information criterion; ICL= Integrated complete-data likelihood criterion.
The overall sample means and conditional response means on the 27 channel variables
used to interpret the five clusters are summarized in Appendix E. The conditional response
means per cluster are also illustrated in Figure 2.10
8 The mclust package uses a definition of BIC that has the opposite sign of the conventional definition: -2 * loglik +
nparams * log(n). Mclust defines BIC as 2 * loglik - nparams * log(n), which denotes that BIC should be maximized instead of minimized (as is usually sought out) (i.e. In this case, the larger the BIC-value, the clearer the evidence for the respective model and its affiliated number of clusters) (Fraley & Raftery, 2003; Scrucca et al., 2016).
9 Two- to five-cluster models were tested iteratively in MPlus to determine which of these models best fit the data.
Each model was evaluated using the BIC, the Akaike information criteria (AIC) (Akaike, 1974) and the Bootstrapped Likelihood Ratio Test (BLRT) (McLachlan & Peel, 2000). The five-cluster model emerged as the best-fitting solution. The conditional response means for the five-cluster model in MPlus were similar, but not identical to those emerging from the original latent profile analysis conducted in R, using the mclust package. This is likely due to differing parameter start values used in the analyses in MPlus and R. Plotted graphically, the conditional response means per cluster resulting from the analysis in MPlus demonstrated an overall cluster-structure which corresponded to the cluster-cluster-structure resulting from the analysis in R.
Figure 2. Distribution of conditional response means on media use variables for the five-cluster solution.11
11 News-channel usage measured on a 7-point scale: 1 = “never,” 2 = “less than once a month,” 3 = “1 to 3 times per month,” 4 = “1 to 3 times per week,” 5 = “4
to 6 times per week,” 6 = “daily,” 7 = “several times a day.” 1 2 3 4 5 6 Me an usa g e
Globally oriented traditionalists: 27.2%/N News junkies: 14.6%/N
Nationally oriented dabblers: 22.8%/N News avoiders: 11.4%/N
The first cluster contains 27.2% of the sample and is the largest cluster. It is characterized
by a focus on national and international news channels, with online editions of national
newspapers and websites of international news providers claiming some of the highest
conditional response means in this cluster. It is also defined by a comparatively high reliance on
traditional media (public service television, radio, and online editions of newspapers). Striking
about this cluster are the low conditional response means on non-serious, soft news channels,
such as tabloids, boulevard magazines on television, and web portals like Yahoo!. Moreover, the
cluster exhibits low levels of new (especially social) media use for news purposes, except for
Facebook. Accordingly, the news repertoire corresponding to this cluster was labeled globally
oriented traditionalists.
The next cluster represents a news junkies repertoire. It is a small cluster, containing only
14.6% of the sample. It is the most diverse cluster out of all five clusters as it exhibits moderate
to high use of all 27 news channels. In fact, with a few exceptions, this cluster displays higher
conditional response means than any other cluster on every news channel.
The third cluster resembles the second cluster in terms of diversity. However, unlike in
the second cluster, conditional response means on the variety of channels represented are
moderate to low. This cluster stands apart from the other clusters for its members’ comparatively
high use of radio. Further, individuals in this cluster rely mostly on public service television,
Facebook, WhatsApp, and interpersonal communication to get their news. Finally, this cluster
can be distinguished by a national and regional orientation: It displays distinctly low conditional
response means on both international television and international news websites. Accordingly,
the news repertoire corresponding to this cluster was labeled nationally oriented dabblers 22.8% of the sample belongs to this cluster.
The penultimate cluster is the smallest cluster, containing 11.4% of the sample. It was
labeled news avoiders because individuals in this cluster exhibit lower conditional response
means than individuals in any other cluster on all news channels except for web portals and
Snapchat. In this cluster, Facebook and interpersonal conversation are relied upon most as news
sources.
The last cluster is the second largest cluster, encompassing 24% of the sample. It is
characterized by high use of social media and interpersonal conversation for news purposes.
Except for YouTube and Twitter, which are used more by the news junkies, all social media
channels are more frequently used in this cluster than in any other cluster. Individuals in this
cluster additionally rely on public service television, the online editions of regional and national
newspapers, and news apps to get their news. Accordingly, the news repertoire corresponding to
this cluster was labeled online news users.
Regarding the first hypothesis, there is evidence of both modality and
channel/content-centered repertoires. Whereas the online news users repertoire is clearly structured around new
(mainly social) media, the nationally oriented dabblers repertoire is structured around national
and regional news content. The globally oriented traditionalists group has a distinct focus on
both a specific modality (traditional media) and a certain type of content/channel (national and
international news). The news junkies and news avoiders repertoires contain both a variety of
content/channels and modalities. These results are in line with H1.
Socio-Demographic Makeup of Repertoires (H2 and H3)
H2 and H3 stated that young adults (ages 20 to 25) would have more diverse and more traditional-media-centered news repertoires than adolescents (ages 16 to 19). To address these
through LPA, were transferred into SPSS Version 22.0 (IBM, 2013) as a new
‘cluster-membership’ variable (i.e. a variable indicating the repertoire which each respondent most probably belonged to). Each repertoire was then investigated according to its socio-demographic
makeup (summarized in Table 2).
Table 2.
Characteristics of Young News Users in Each Repertoire Globally oriented traditionalists (N = 43) News junkies (N = 23) Nationally oriented dabblers (N = 36) News avoiders (N = 18) Online news users (N = 38) Gender Male 34.9 52.2 22.2 22.2 23.7 Female 62.8 47.8 77.8 77.8 73.7
Prefer not to say 2.3 0 0 0 2.6
χ2(8, N = 158) = 10.02, p = .264 Age M (SD) 22.42 (2.54) 22.74 (2.56) 22.14 (2.67) 22.89 (2.45) 21.76 (3.06) 16-19 16.3 17.4 19.4 5.6 23.7 20-25 83.7 82.6 80.6 94.4 76.3 χ2(4, N = 158) = 2.89, p = .576 Education No schooling completed 0 0 0 0 2.6
Some high school, no diploma 2.3 0 2.8 5.6 5.3
High school 23.3 34.8 36.1 27.8 28.9 Vocational training 4.7 4.3 16.7 5.6 7.9 Bachelor’s degree 44.2 43.5 33.3 44.4 44.7 Master’s degree 25.6 17.4 11.1 16.7 10.5 χ2(20, N = 158) = 14.97, p = .778 Political orientation M (SD) 3.93 (1.71) 3.57 (1.88) 3.92 (1.81) 3.78 (0.94) 3.47 (1.75) F (4, 153) = .53, p = .717
Note. Values are in percentages, except for where indicated otherwise.
In terms of gender, the news junkies and the globally oriented traditionalists have the
most even distribution. The other repertoires are female-dominated. The globally oriented
traditionalist group is the most educated of all repertoires, with 69.9% of its members possessing
a bachelor’s degree or higher-level qualification. It is followed by the news avoiders, news
junkies, online news users, and finally, by the nationally oriented dabblers. In political
orientation, the five repertoires do not exhibit significant differences; all repertoire groups are
rather left-leaning.
Regarding age, there are no significant differences in mean age between the five
repertoires. However, when the distributions of young adults and adolescents across the
repertoires are examined, some nuances in age-difference can be discerned: The online news
users group contains the largest percentage of adolescents (23.7%) out of all five repertoires,
lending some support to the hypothesis that the news repertoires of young adults feature
traditional media more prominently than those of adolescents (H3). The comparatively high
percentage of young adult globally oriented traditionalists (83.7%) a repertoire with a strong
focus on traditional media corroborates this tendency.
In terms of repertoire-diversity (H2), the most diverse repertoire, i.e. the news junkies,
features a comparatively high percentage of 20-to-25-year-olds (82.6%), substantiating the
hypothesis that young adults have more diverse news repertoires than adolescents. However, the
remarkably large amount of 20-to-25-year-olds (94.4%) in the news avoiders repertoire
(arguably the least diverse repertoire in terms of actual time spent attending to the news) goes
against this finding. These results partially confirm H2 and H3.
Repertoires and Political Knowledge- and Efficacy (H4a,H4b, H5 and RQ1)
efficacy, three one-way ANCOVAs were conducted using political knowledge, internal-, and
external political efficacy as dependent variables respectively and repertoire group as the
independent variable, while controlling for age, gender, education and political orientation.
Political knowledge. Results of the first ANCOVA indicated that the repertoires differ
significantly in terms of political knowledge, whilst controlling for age, gender, education and
political orientation (F (4, 149) = 5.94, p < .001, partial η2 = .14) (see Appendix F for summary
table). Post-hoc Bonferroni tests (visualized in Figure 3) revealed significant differences between
the globally oriented traditionalists (Estimated Marginal Mean (EMM) = 2.98) and: (1) the
nationally oriented dabblers (EMM = 2.09), (2) the news avoiders (EMM = 1.88), (3) the online news users (EMM = 2.22).
These results are in line with H5, which states that news repertoires prominently featuring
traditional media are more positively associated with political knowledge than those that
prominently feature new media. Indeed, the globally oriented traditionalists repertoire, for which
the highest level of political knowledge is estimated, features traditional media most prominently
out of all five repertoires and also exhibits significantly higher levels of political knowledge than
the new-media-centered online news users repertoire.
These results do not as clearly confirm H4a, as there are no significant differences in
political knowledge between the most diverse repertoire (news junkies) and all other repertoires.
Political efficacy. H4b states that more diverse news repertoires will be more positively associated with political efficacy than less diverse repertoires. RQ1 asks to what extent news
repertoires which prominently feature traditional media, versus those which prominently feature
new media, are differentially associated with political efficacy. The results of the second and
The second ANCOVA demonstrated that the repertoires differ significantly in terms of
internal political efficacy, whilst controlling for age, gender, education and political orientation
(F (4, 149) = 13.41, p < .001, partial η2 = .27). A number of significant differences emerged from
post-hoc tests (Figure 4): The news avoiders (EMM = 2.46) repertoire arises as the repertoire
with significantly lower levels of internal political efficacy than any other repertoire. Further, the
globally oriented traditionalists (EMM = 3.83) exhibit significantly higher levels of internal
political efficacy than the online news users (EMM = 3.20), the nationally oriented dabblers
(EMM = 3.18), and the news avoiders.
The third ANCOVA indicated that the repertoires also differ significantly in terms of
external political efficacy, whilst controlling for age, gender, education and political orientation
(F (4, 149) = 2.60, p = .039, partial η2 = .07). The highest level of external political efficacy is
estimated for the online news users (EMM = 3.30) as well as the globally oriented traditionalists
(EMM = 3.30). The lowest level is estimated for the news avoiders (EMM = 2.74). However,
post-hoc Bonferroni comparisons revealed only two significant differences between repertoires
(see Figure 5): The news avoiders differ significantly in external political efficacy from both the
globally oriented traditionalists and the online news users.
Taken together, these results partially confirm H4b. In terms of internal political efficacy,
the least diverse repertoire (news avoiders) exhibits significantly lower levels of internal efficacy
than all other repertoires, while the considerably diverse globally oriented traditionalists group
displays significantly higher levels of internal political efficacy than all other, less diverse
repertoires. When it comes to external political efficacy, the news avoiders display significantly
lower levels than the globally oriented traditionalists, but the superiority of more diverse
significantly surpass the news avoiders is the moderately diverse online news users group.
Regarding RQ1, the news repertoires which prominently feature traditional media (i.e. the
globally oriented traditionalists and the news junkies) exhibit significantly higher levels of
internal political efficacy than those which feature less traditional media (i.e. online news users).
This does not hold for external political efficacy, which is highest for both the
traditional-media-centric globally oriented traditionalists repertoire and the new-media-focused online news users
repertoire.
Figure 3. Estimated marginal means (EMM) and EMM-differences (Bonferroni) for political
knowledge by repertoire group.
1 2 3 4 5 (1) Globally oriented traditionalists (2) News junkies (3) Nationally oriented dabblers (4) News avoiders (5) Online news users Est im at ed m ar gi nal m eans 0 1 2 3 4 (1) Globally oriented traditionalists (2) News junkies (3) Nationally oriented dabblers (4) News avoiders (5) Online news users Est im at ed m ar gi nal m eans
Figure 4. Estimated marginal means (EMM) and EMM-differences (Bonferroni) for internal
Figure 5. Estimated marginal means (EMM) and EMM-differences (Bonferroni) for external
political efficacy by repertoire group.
Conclusion and Discussion
This study sought to identify young people’s news repertoires and to examine the associations between these repertoires and political knowledge and -efficacy.
Using an exhaustive list of news channels, five news repertoires were identified: Globally
oriented traditionalists (centered around national and international content and traditional
media), news junkies (characterized by high usage of all news sources), nationally oriented
dabblers (moderate news users focused mainly on national content via a variety of modalities), news avoiders (low news consumers across all sources), and online news users (focused on
online news sources and social media).
The existence of news repertoires which predominantly feature traditional media
(globally oriented traditionalists) and those which are characterized by an overall avoidance of
(news avoiders), or a firm enthusiasm for all varieties of news (news junkies), has been found in
previous news repertoire studies using largely adults (e.g., Edgerly, 2015; Kim, 2014). Equally in
line with previous repertoire studies (e.g., Hasebrink & Schmidt, 2012; Kim, 2014), the news
1 2 3 4 5 (1) Globally oriented traditionalists (2) News junkies (3) Nationally oriented dabblers (4) News avoiders (5) Online news users Est im at ed m ar gi nal m eans
repertoires identified were structured around both modality and content/channel.
The identification of two news repertoires with a clear focus on geographically-oriented
news content/channels (globally oriented traditionalists and nationally oriented dabblers)
emerged as a comparatively novel finding.12 It implies that some young media users pattern their
news use around the geographical areas most pertinent to them. In this regard, an interesting line
of future research would be to investigate to what extent sociodemographic indicators not
included in the present study (e.g., residence in a rural versus urban area, living in a binational
household) are associated with these geographically-oriented news repertoires.
A further finding which has no precedent in older repertoire studies, is the emergence of a
social-media-dominated news repertoire (online news users). While older studies have identified
similar digitally-centered news repertoires (e.g., Schrøder & Kobbernagel, 2010; Swart et al.,
2016), none of them used a comprehensive list of modern-day social media to identify these
repertoires, as was done in the present study, which included all major contemporary social
media outlets.13 The inclusion of these social media outlets ultimately resulted in an entire news
repertoire focused primarily around these channels. This finding suggests that certain young
people may be moving toward a completely social-media based mode of news consumption. It
also opens up promising avenues for future research into the particular character of news
encountered within this social-media-centered repertoire (i.e. piecemeal, ‘stumbled-upon’ news
or actual, in-depth news?).
Furthermore, the discovery of a traditional-media-centered repertoire (globally oriented
traditionalists) implies that new media has not, as some researchers have previously posited
12 Only one other news repertoire study (Swart et al., 2016) has identified geographically-oriented news repertoires.
However, unlike the present study, it used a qualitative approach to determine these news repertoires.
13 The social media channels included in the present study were: Facebook, Instagram, Twitter, Snapchat, YouTube,
(e.g., Gaskins & Jerit, 2012; Patterson, 2007), replaced traditional media in young people’s news
consumption habits (at least not for all youth).
No significant age-differences were identified between the five repertoires. This suggests
that the age-effect may not condition young people’s news consumption as strongly as initially
thought: Adolescents seem to be equally interested and diverse news users as young adults. On
the other hand, the fact that the online news users represented the youngest-leaning repertoire,
implies that a generation effect may in fact exist (i.e. Young adults use traditional media more
frequently than adolescents, who preferably frequent online/new media). However, this may also
point to another type of age-effect, whereby adolescents start out using largely online/social
media for news consumption and transition into using more ‘serious,’ traditional media for news-purposes as they mature.
Moreover, significant differences in political knowledge emerged between the
repertoires. The globally oriented traditionalists exhibited significantly higher levels of political
knowledge than all other repertoires save for the news junkies, strengthening the often-made
argument that traditional media promote the acquisition of political knowledge more effectively
than new media (e.g., Dimitrova et al., 2011). This finding also provides an insight into the
association between repertoire-diversity and political knowledge: The fact that the most diverse
repertoire (i.e. news junkies) did not exhibit significantly higher levels of political knowledge
than any of the other (less diverse) repertoires, suggests that the acquisition of political
knowledge may have less to do with the diversity of a young person’s news repertoire and more
with the extent to which it features traditional media.
Significant differences between the repertoires also emerged for internal political
other repertoires save for the news junkies, signifying the central role of traditional-media-centric
repertoires in the promotion of internal efficacy. Importantly, the non-diverse news avoiders
repertoire exhibited significantly lower levels of internal efficacy than all other repertoires,
indicating that repertoire-diversity is a further important factor in the promotion of internal
political efficacy.
Interestingly, results for external efficacy presented a different picture: Both the globally
oriented traditionalists and the online news users showed significantly higher levels of external
efficacy than the news avoiders, suggesting that repertoire-diversity is not essential for its
promotion and that instead, both a focus on traditional (i.e. globally oriented traditionalists) and
on online/new media (i.e. online news users) can stimulate external efficacy. This ability of
new-media-centric news repertoires to promote external political efficacy presumably stems from the
inherent communicativeness of new media technologies (especially social media): It helps build
closer relationships between young people and their elected representatives through online
interactions such as conversations on Twitter, or Reddit “Ask Me Anything” sessions, effectively giving them the feeling that politicians are responsive to their ideas, opinions and worries (Zhou
& Pinkleton, 2012).
Overall, the differences in political knowledge and -efficacy found between the
repertoires suggest that the prominence of traditional media in news repertoires is a central factor
in the cultivation of both political knowledge and -efficacy, more so than the extent to which a
news repertoire is diverse. This insight has wide-ranging societal implications. For example, it
can provide valuable guidance for structuring high school civics curricula in such a way as to
best promote young students’ capacities to become politically knowledgeable and -efficacious citizens, who are well-equipped to participate in the democratic process.
Despite its valuable findings, this study has several limitations that must be
acknowledged. Firstly, as with most survey-research, the cross-sectional nature of the study does
not allow any absolute statements about causality. This is important when considering the
relationships found between news repertoires and political knowledge and -efficacy, which may
very well be reciprocal or run in the opposite direction of what has been assumed throughout the
study.14
Secondly, the study relied on self-reports to establish respondents’ media consumption
habits. Self-reported media use has frequently come under fire for being inaccurate (e.g., Prior,
2009). While more exact behavioral measures of news media use were not feasible for this study,
the use of such methods to identify (news) media repertoires presents an interesting avenue for
future research, with some initial studies already using approaches such as media-use tracking
(Yuan & Webster, 2006) and participant observation (Taneja et al., 2012) successfully.
Thirdly, the sample used for this study was a convenience sample, in which females and
young adults (20 to 25 years) were overrepresented, making generalizability to the wider
population of German youth difficult. Similarly, the present study identified the news repertoires
of German youth at one specific point in time only, making it difficult to determine to what
extent these repertoires can be generalized across time. However, the continuous transformation
of media technologies and consumption patterns implies that any study of media use will be
strongly time-bound, making “even the highest quality sample surveys become dated quickly and
thereby limited in generalizability…” (Althaus & Tewksbury, 2000, p. 22). Moreover, this shortcoming opens up avenues for future research to test the presently identified repertoires on
14 For example, the globally oriented traditionalists repertoire may not so much promote political knowledge as it is
other (cross-national) samples,15 as well as longitudinally, across different time periods.
Despite its limitations, this study has provided important insights into the patterns by
which young people structure their news exposure experiences in the modern, high-choice media
environment. It is one of very few existing studies to use a repertoire approach to study young
people’s news use and the first to examine the associations of these repertoires with political knowledge and -efficacy. It presents a springboard for future studies to investigate potential
predictors (e.g., socioeconomic status), as well as further democratic implications (e.g., political
participation) of young people’s news repertoires. Finally, this study has elucidated the
intricacies of young people’s news consumption habits in a far more informative way than what can be gleaned from standard figures on mean youth news use or general news audience
statistics.
15 Testing the presently identified news repertoires on other samples (especially on a more representative sample of
the German population of 16-to-25-year-olds) is especially important because the highly educated and
internationally-oriented sample used in the present study may have resulted in a disproportionately large group of
References
AGJ. (2017). Deutscher kinder- und jugend(-hilfe) monitor 2017. Retrieved 03.05.2017, from
https://www.agj.de/fileadmin/files/arbeitsfelder/Monitor_Dt_Kinder-jugendhilfe_FINAL.pdf
Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on
Automatic Control, 19(6), 716-723. doi: 10.1109/TAC.1974.1100705
Althaus, S. L., & Tewksbury, D. (2000). Patterns of internet and traditional news media use in a
networked community. Political Communication, 17(1), 21-45. doi:
10.1080/105846000198495
Amnå, E., Munck, I., & Zetterberg, P. (2004, April). Meaningful participation? Political efficacy
of adolescents in 24 countries. Emerging repertoires of political action: Toward a systemic study of post-conventional forms of participation. Paper presented at the ECPR
Joint Sessions of Workshops, Uppsala, Sweden.
https://ecpr.eu/Filestore/PaperProposal/e8a5f984-6738-4c7b-bebf-f2e385562b0f.pdf
Antunovic, D., Parsons, P., & Cooke, T. R. (2016). ‘Checking’ and googling: Stages of news
consumption among young adults. Journalism, 17(0), 1-17. doi:
10.1177/1464884916663625
Bakker, T. P., & de Vreese, C. H. (2011). Good news for the future? Young people, internet use,
and political participation. Communication Research, 38(4), 451-470. doi:
10.1177/0093650210381738
Baumgartner, J. C., & Morris, J. S. (2010). MyFaceTube politics. Social Science Computer
Bennett, W. L. (2008). Changing citizenship in the digital age. In W. L. Bennett (Ed.), Civic life
online: Learning how digital media can engage youth (pp. 1-24). Cambridge, MA: The
MIT Press.
Bennett, W. L. (2012). The personalization of politics. The ANNALS of the American Academy of
Political and Social Science, 644(1), 20-39. doi: 10.1177/0002716212451428
Biernacki, C., Celeux, G., & Govaert, G. (2000). Assessing a mixture model for clustering with
the integrated completed likelihood. IEEE Transactions on Pattern Analysis and Machine
Intelligence, 22(7), 719-725. doi: 10.1109/34.865189
Bovill, M., & Livingstone, S. (2001). Children and their changing media environment: A
European comparative study. Mahwah, NJ: Lawrence Erlbaum Associates.
bpb. (2012). Bevölkerung nach altersgruppen und geschlecht. Retrieved 03.05.2017, from
http://www.bpb.de/nachschlagen/zahlen-und-fakten/soziale-situation-in-deutschland/61538/altersgruppen
Carpini, M. X. D., & Keeter, S. (1993). Measuring political knowledge: Putting first things first.
American Journal of Political Science, 37(4), 1179-1206. doi: 10.2307/2111549
Carpini, M. X. D., & Keeter, S. (1997). What Americans know about politics and why it matters.
New Haven, CT: Yale University Press.
Casero-Ripollés, A. (2012). Beyond newspapers: News consumption among young people in the
digital era. Comunicar, 20(39), 151-158. doi: 10.3916/C39-2012-03-05
Choi, J. (2016). Why do people use news differently on SNSs? An investigation of the role of
motivations, media repertoires, and technology cluster on citizens' news-related activities.
Computers in Human Behavior, 54, 249-256. doi:
Craft, S., Ashley, S., & Maksl, A. (2016). Elements of news literacy: A focus group study of how
teenagers define news and why they consume it. Electronic News, 10(3), 143-160. doi:
10.1177/1931243116656716
Craig, S. C., & Maggiotto, M. A. (1982). Measuring political efficacy. Political Methodology,
8(3), 85-109. doi: http://www.jstor.org/stable/25791157
de Vreese, C. H., & Boomgaarden, H. (2006). News, political knowledge and participation: The
differential effects of news media exposure on political knowledge and participation.
Acta Politica, 41(4), 317-341. doi: 10.1057/palgrave.ap.5500164
Diddi, A., & LaRose, R. (2006). Getting hooked on news: Uses and gratifications and the
formation of news habits among college students in an internet environment. Journal of
Broadcasting & Electronic Media, 50(2), 193-210. doi: 10.1207/s15506878jobem5002_2
Dimitrova, D. V., Shehata, A., Strömbäck, J., & Nord, L. W. (2011). The effects of digital media
on political knowledge and participation in election campaigns. Communication
Research, 41(1), 95-118. doi: 10.1177/0093650211426004
Dutta-Bergman, M. J. (2004). Complementarity in consumption of news types across traditional
and new media. Journal of Broadcasting & Electronic Media, 48(1), 41-60. doi:
10.1207/s15506878jobem4801_3
Edgerly, S. (2015). Red media, blue media, and purple media: News repertoires in the colorful
media landscape. Journal of Broadcasting & Electronic Media, 59(1), 1-21. doi:
10.1080/08838151.2014.998220
Edgerly, S., Thorson, K., Thorson, E., Vraga, E. K., & Bode, L. (2017). Do parents still model
news consumption? Socializing news use among adolescents in a multi-device world.
Eshghi, A., Haughton, D., Legrand, P., Skaletsky, M., & Woolford, S. (2011). Identifying
groups: A comparison of methodologies. Journal of Data Science, 9, 271-291.
Fraley, C., & Raftery, A. E. (1998). How many clusters? Which custering method? Answers via
model-based cluster analysis. The Computer Journal, 41(8), 578-588. doi:
10.1093/comjnl/41.8.578
Fraley, C., & Raftery, A. E. (2003). Enhanced model-based clustering, density estimation, and
discriminant analysis software: MCLUST. Journal of Classification, 20(2), 263-286. doi:
10.1007/s00357-003-0015-3
Gaskins, B., & Jerit, J. (2012). Internet news. The International Journal of Press/Politics, 17(2),
190-213. doi: doi:10.1177/1940161211434640
Gil de Zúñiga, H., Jung, N., & Valenzuela, S. (2012). Social media use for news and individuals'
social capital, civic engagement and political participation. Journal of
Computer-Mediated Communication, 17(3), 319-336. doi: 10.1111/j.1083-6101.2012.01574.x
Gil de Zúñiga, H., Puig-I-Abril, E., & Rojas, H. (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. doi: 10.1177/1461444809102960
Hagenaars, J. A., & McCutcheon, A. L. (2002). Applied latent class analysis. Cambridge, UK:
Cambridge University Press.
Hao, X., Wen, N., & George, C. (2014). News consumption and political and civic engagement
among young people. Journal of Youth Studies, 17(9), 1221-1238. doi:
Hasebrink, U., & Domeyer, H. (2012). Media repertoires as patterns of behaviour and as
meaningful practices: A multimethod approach to media use in converging media
environments. Participations: Journal of Audience & Reception Studies, 9(2), 757-779.
Retrieved from http://www.participations.org/Volume%209/Issue%202/contents.htm
Hasebrink, U., Jensen, K. B., van den Bulck, H., Hölig, S., & Maeseele, P. (2015). Changing
patterns of media use across cultures: A challenge for longitudinal research. Journal of
Communication, 9, 435–457. Retrieved from
http://ijoc.org/index.php/ijoc/article/view/3452/1303
Hasebrink, U., & Popp, J. (2006). Media repertoires as a result of selective media use: A
conceptual approach to the analysis of patterns of exposure. Communications, 31(3),
369-387. doi: 10.1515/COMMUN.2006.023
Hasebrink, U., & Schmidt, J. H. (2012). Informationsrepertoires der deutschen Bevölkerung.
Konzept fur eine regelmäßig durchzufuhrende bevölkerungsrepräsentative befragung im rahmen des vorhabens “Erfassung und darstellung der medien und meinungsvielfalt in Deutschland" (Vol. 24). Hamburg: Hans-Bredow-Institut.
Heeter, C. (1985). Program selection with abundance of choice. Human Communication
Research, 12(1), 126-152. doi: 10.1111/j.1468-2958.1985.tb00070.x
Huang, E. (2009). The causes of youths' low news consumption and strategies for making youths
happy news consumers. Convergence, 15(1), 105-122. doi: 10.1177/1354856508097021
IBM. (2013). IBM SPSS Statistics for Windows (Version 22.0). Armonk, NY: IBM Corp.
Kaid, L. L., McKinney, M. S., & Tedesco, J. C. (2007). Political information efficacy and young
voters. American Behavioral Scientist, 50(9), 1093-1111. doi:
Kim, S. J. (2014). A repertoire approach to cross-platform media use behavior. New Media &
Society, 18(3), 353-372. doi: 10.1177/1461444814543162
Ksiazek, T. B., Malthouse, E. C., & Webster, J. G. (2010). News-seekers and avoiders:
Exploring patterns of total news consumption across media and the relationship to civic
participation. Journal of Broadcasting & Electronic Media, 54(4), 551-568. doi:
10.1080/08838151.2010.519808
Lai, C.-H., & Tang, T. (2015). Understanding local news consumption and community
participation via the lens of information repertoires and media multiplexity. Mass
Communication and Society, 18(3), 325-349. doi: 10.1080/15205436.2014.995768
LaRose, R. (2010). The problem of media habits. Communication Theory, 20(2), 194-222. doi:
10.1111/j.1468-2885.2010.01360.x
Lee, H., & Yang, J. (2014). Political knowledge gaps among news consumers with different
news media repertoires across multiple platforms. International Journal of
Communication, 8, 21. doi: http://ijoc.org/index.php/ijoc/article/view/2455
Loader, B. D., Vromen, A., & Xenos, M. A. (2014). The networked young citizen: Social media,
political participation and civic engagement. Information, Communication & Society,
17(2), 143-150. doi: 10.1080/1369118X.2013.871571
Magidson, J. M., & Vermunt, J. K. (2002). Latent class models for clustering: A comparison
with k-means. Canadian Journal of Marketing Research, 20(1), 36-43.
McLachlan, G., & Peel, D. (2000). Assessing the number of components in mixture models
Finite mixture models (pp. 175-220). New York, NY: John Wiley & Sons, Inc.
Miller, W. E., Miller, A. H., & Schneider, E. J. (1980). American national election studies data
