Available via license: CC BY-NC 4.0
Content may be subject to copyright.
https://doi.org/10.1177/1464884919899313
Journalism
1 –19
© The Author(s) 2020
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1464884919899313
journals.sagepub.com/home/jou
Newsworthiness and story
prominence: How the presence
of news factors relates to
upfront position and length
of news stories
Mark Boukes , Natalie P Jones
and Rens Vliegenthart
University of Amsterdam, The Netherlands
Abstract
The presence of news factors in journalistic products has been abundantly researched,
but investigations into their actual impact on the news production process are scarce.
This study provides a large-scale analysis of why news factors matter: Whether,
how, and which news factors affect the prominence of news items and does this
differ per outlet type? A manual content analysis of print, online, and television
news demonstrates that a larger total number of news factors in a story positively
predict an item’s length and likelihood of front-page publication or likelihood of
being a newscast’s opening item. News factors ‘conflict’ and ‘eliteness’ have the
strongest impact, mixed evidence was found for ‘proximity’ and ‘personification’,
whereas relationships with ‘negativity’, ‘influence and relevance’, and ‘continuity’
were mostly insignificant. Fewer differences than expected emerged between outlet
types (popular vs quality press). Especially for television news, outlet type (public vs
commercial broadcaster) hardly mattered.
Keywords
Content analysis, news factors, news value theory, newsworthiness, prominence,
popular vs quality, public vs commercial
Corresponding author:
Mark Boukes, Amsterdam School of Communication Research (ASCoR), University of Amsterdam, Nieuwe
Achtergracht 166, 1018 WV Amsterdam, The Netherlands.
Email: markboukes@gmail.com
899313JOU0010.1177/1464884919899313JournalismBoukes et al.
research-article2020
Original Article
2 Journalism 00(0)
Over the years, research on news factors has resulted in increasingly sophisticated
taxonomies (from Galtung and Ruge, 1965 to Harcup and O’Neill, 2017). Yet,
important questions about why these news factors concretely matter for news pro-
duction have largely remained unaddressed. Prior research evidenced the presence
of various news factors in journalistic coverage. Insights from within the news-
room, moreover, demonstrated how journalists (Jacobs and Tobback, 2013) and
their sub-editors (Vandendaele, 2018) consciously apply these news factors to
make the most appealing news product. Still little is known, though, about their
consequences for journalistic products and a systematic, large-scale assessment
was lacking.
As Cotter (2010) writes, news factors play a role ‘from the beginning to the end of the
reporting and editing process’ (p. 74), and thus do not only involve the selection of news,
but also decisions about the prominence given to a story. After all, prominence reflects
the degree of importance given by journalists and/or editors (newsworthiness, see Schulz,
1982). This prominence can be operationalized by a news story’s length and placement
(Cotter, 2010; Elorza, 2014; Fico and Freedman, 2001). This study not only compares
the general impact of particular news factors but also whether their impact differs
between popular and quality press,1 and between newscasts of public and commercial
TV channels.
We investigate the impact of news factors within the context of Dutch economic
news. The topic of the economy has the advantage that is found essentially within any
section of a news outlet (domestic and foreign news, culture or sports) – making it one of
the most general news topics. Moreover, a massive increase in economic and business
news has occurred, with almost every platform incorporating economic news to a large
degree (Lee and Baek, 2018). As such, economic news provides a suitable sample for
this study’s purpose (Tumber, 1993).
Newsworthiness and news factors
The study of journalistic news factors can be approached from the following two
theoretical perspectives: a functional and causal model (Staab, 1990). In the func-
tional model, an event is not newsworthy in itself, but is accorded its newsworthiness
by discursively ascribing news factors through language, image, and typography to
sell an event to an audience as news (Bednarek and Caple, 2014). Conceptually,
news factors are assumed to be qualities of a text rather than inherent characteristics
of an event itself, and are applied by the media to heighten the legitimacy of an event
becoming news (Bednarek and Caple, 2014). In the causal model, by contrast, news
factors are inherent qualities of a story that determine whether and how journalists
treat the story.
In either model, the assumption is that the more news factors a story contains, the
more newsworthy it is considered and the higher the likelihood for the event to reach
a prominent publication. The consequences of news factors – whether ascribed by
journalists or as objective event criteria – for the decisions that journalists take
regarding newsworthiness, however, remain not only an assumption to be tested
Boukes et al. 3
(Kepplinger and Ehmig, 2006), but is also at the core of news value theory (Galtung
and Ruge, 1965; Harcup and O’Neill, 2001). Empirically valid tests, with news
selection as the outcome variable, are scarce because it requires a comparison of
extra-media and intra-media events (i.e. what is published and what is not). Also in
the current study, we can not focus on what is published (or not), but instead analyze
the prominence of a news item itself. We assume that selection and prominence are
to a considerable extent driven by the same mechanisms (see also Cotter, 2010);
however, it should be noted that individual journalists are more in charge of topic
selection, whereas decisions about the composition of an outlet (i.e. prominence) are
often taken on higher editorial levels. Nevertheless, the factors determining the
newsworthiness of a story are likely to be the same ones that affect its prominence.
In a study on politicians in Swiss newspapers, for example, Tresch (2009) found that
presence and prominence of actors are to a considerable degree driven by the same
mechanisms and are ‘relatively similar’ (p. 84). By operationalizing newsworthiness
of a news item as its relative prominence in an outlet (see Fico and Freedman, 2001),
this study still allows for the examination of the relationship between news factors’
presence and journalistic decisions of newsworthiness.
Prominence has the advantage that it allows for a transparent and objective measure-
ment, which is virtually unfeasible regarding the original gatekeeping processes of story
selection. Specifically, prominence is operationalized as a function of both story length
and story position in an outlet (Elorza, 2014). In essence, the more newsworthy a news
item is considered by media workers – because it contains more news factors – the more
prominence it should be assigned and the longer and earlier the article should appear
within a news product (Schulz, 1982). Following the theoretical idea behind newsworthi-
ness (Staab, 1990), we expect the following:
H1: The more news factors a news item contains, the more prominence it will be
given in terms of (a) story length and (b) story position.
Identification of news factors
Ever since Galtung and Ruge’s (1965) cornerstone research, clear standards for which
news factors to include in empirical research have remained absent. Harcup and O’Neill
(2017) even revised their own 2001 revision of Galtung and Ruge’s taxonomy of news
values, as it transpired that some of their newly defined categories were too broad when
subject to empirical analysis. We adopt the framework of news factors that have repeat-
edly been found to be relevant in the North-European region (Eilders, 2006). This frame-
work largely overlaps with the news factors identified in a research project across 63
countries (Masterton, 2005) and corresponds with the criteria for newsworthiness expli-
cated by journalism textbooks over the past decades (Parks, 2019).
Specifically, Eilders (2006) found seven news factors to continually influence journalis-
tic judgments of newsworthiness: negativity (damage; the negative aspects of an event or
issue), continuity (frequency; having been in the news previously), proximity (cultural, geo-
graphical, and/or economic ‘nearness’), eliteness (presence of individuals, organizations, or
4 Journalism 00(0)
nations of elite status involved in an event), influence and relevance (the significance of an
event in terms of its effects and/or consequences), personification (inclusion of the personal
in an event, such as eyewitness reports), and conflict (presence of confrontation and/or con-
troversy). Whereas, the presence of these news factors has repeatedly been investigated,
literature is still lacking an examination of which specific news factors actually relate to the
prominence of news items. Few stories will contain all news factors, as stories are inherently
bounded and constrained in terms of length. Because some news factors require more space
(e.g. personification) than others (e.g. continuity) do, the question of which news factors are
most important in determining prominence becomes a highly relevant one – especially when
story length and story position are considered separately. Without much literature to formu-
late clear-cut hypotheses, we investigate whether all news factors or only some of these
relate to story prominence:
RQ1: Which news factors relate to news item prominence in terms of (a) story length
and (b) story position?
News factors by outlet type
News factors and news values are the two components that comprise newsworthiness
(Kepplinger and Ehmig, 2006). As aforementioned, news factors are inherent or ascribed
event characteristics influencing how journalists evaluate and select a story for publica-
tion (Staab, 1990). News values, in contrast, are the valuations by journalists regarding
the relevance of specific news factors. As such, news factors themselves do not deter-
mine the newsworthiness of a story; a journalist must assign them sufficient value. This
study elucidates whether differences exist between outlet types in the value that they
assign to the specific news factors.
The organizational structure, commercial pressures, and motivations differing
between news media may contribute to a particular construction of newsworthiness
(Bednarek and Caple, 2014) by giving more value to one news factor than to another
(Allern, 2002). Growing scholarly concern about the ‘tabloidization’ of news, in which
tabloid news factors (i.e. personalization and negativity) would be increasingly adopted
by traditional media at the expense of journalistic quality, warrants research into the
validity of such claims (Barnett and Seymour, 1999).
Concretely, we compare quality versus popular news media and public versus com-
mercial broadcasting newscasts. It is important to emphasize that we are interested in the
relative importance of different news factors across types of media. One can anticipate
that the specific news factors contribute to prominence across the board, but that for dif-
ferent outlet types, certain news factors are considered to be of more value and, thus, to
contribute more strongly to stories’ newsworthiness.
News factors relevant for the quality news media
Readers of quality press are often referred to as ‘elite audiences’; educated consumers
seeking to be informed about ‘serious news’ rather than about the lighter fare (Mitchell
and Holcomb, 2016). While the news audience generally declines, readership of quality
Boukes et al. 5
news has remained relatively steady due to the stable news interest of their audience
niche. Feeling less commercial pressure from declining audiences, quality press are
allowed greater authority over their story construction than popular news media
(Strömbäck et al., 2012).
Quality press journalists frequently emphasize the importance of objective reporting
(Skovsgaard, 2014). Through actively including a range of perspectives and commentary
on an issue, journalists demonstrate their adherence to this objectivity norm. In the litera-
ture (i.e. most prominently by Semetko and Valkenburg, 2000), this has been broadly
defined as the factor of ‘conflict’, which represents a journalistic style of presenting
issues with opposing and clashing views. Conflict, thus, provides a method of compre-
hensive reporting, as including and comparing contrasting positions can present addi-
tional insights that advance audience knowledge (Bartholomé et al., 2015). As such, the
broadly defined news factor of ‘conflict’ may be ascribed more value in quality press
outlets than in the popular press.
Journalists of quality press also regularly delve beneath the surface and approach
stories with a broader political perspective by narrating stories of social significance and
emphasizing the potential societal consequences (Reinemann et al., 2012).2 In fact, com-
mercial news factors such as the episodic focus on personalization and negativity are
often said to displace relevance in popular news outlets (Skovsgaard, 2014). Thus, jour-
nalistic decisions about prominence are likely to be guided more strongly by influence
and relevance as a news factor in quality news outlets compared to popular news outlets.
Prominent stories on the same topic in quality news outlets, therefore, are more likely to
highlight the societal-wide impact of a story compared to a more factual summary or
personified coverage of the story in popular news outlets.
Involvement of elite actors is often an important news factor. While the information
these sources provide may shape the newsworthiness of a story, the presence of these
sources themselves is already considered newsworthy in itself (Strömbäck et al., 2012).
After all, their involvement evokes the impression that an issue must be societally impor-
tant, and requires the attention of journalists who perceive themselves as societal ‘watch-
dogs’ (Skovsgaard, 2014). As quality news media seek to represent a more encompassing
societally relevant story and elaborate analysis, they will have a higher likelihood of
prominently publishing a news item when elite individuals, organizations, or institutions
are present (Caple and Bednarek, 2013).3
Finally, we expect the news factor of continuity to especially pertain to quality news
media. This news factor refers to a journalistic style that stands opposite to fragmented
or incident-based reporting that lacks context and/or interpretation (see Mothes et al.,
2019). Continuity instead refers to a form of journalism in which topics are covered in
greater depth by journalists who have the time to specialize in an issue and give interpre-
tation. Alternatively, as Harcup and O’Neill (2001: 263) write, ‘even if its amplitude has
been greatly reduced’, journalists may follow up on an issue to dig deeper and further
explain it to its audience. Earlier coverage will cause audience awareness of a topic, and
readers may expect the quality outlets to provide more prominent coverage of this topic
to further expand their knowledge (Eilders, 2006). This is not to say that those popular
outlets do not provide any continuity; actually, Mothes et al. (2019) find that non-frag-
mented news is particularly attractive to readers with lower internal efficacy. We rather
6 Journalism 00(0)
contend that the resources and incentives that those outlets have to produce follow-up
coverage are more limited and therefore prioritize covering the ‘breaking news’ of the
day. Based on this theoretical background, the following sub-hypothesis is formulated:
H2: News factors (a) conflict, (b) influence and relevance, (c) eliteness, and (d) con-
tinuity positively relate to story prominence more strongly for quality press compared
to popular press.
News factors relevant for the popular news media
Whereas quality news outlets target elite audiences, the popular press first-and-foremost
strives to appeal to the broadest possible audience (Skovsgaard, 2014). Seemingly con-
tradictory, however, they suffer the most from decreases in news consumption among the
general public. This instills a continuous commercial pressure that can cause popular
news journalists to ascribe increased value to a set of commercial news factors that
would appeal to a large target audience (Allern, 2002). First, it is imperative that popular
news media simplify their stories; so, stories are easily digestible to the average citizen
(Allern, 2002). Personified content offers closeness to as well as identification with the
reader and is more easily comprehensible for audiences with low news interest. Hence,
it may especially appeal to the audience typical of popular news (Mothes et al., 2019),
thus incentivizing popular news media to include citizens’ private experiences by report-
ing events with a personalized viewpoint (Skovsgaard, 2014). Again, this is in compara-
tive perspective – also quality newspapers will have those incentives, but arguably to a
lesser degree.
To target the broadest possible audience, popular news is expected to put a premium
on the news factor of negativity. Curran et al. (1980) illustrate that a personal emphasis
and sensationalized negative elements in news make it more accessible to non-elite audi-
ences. Finally, the news factor proximity aids in journalists’ aim to appeal to the masses:
Proximity increases audience engagement (Trilling et al., 2017), and as such the likeli-
hood that popular news media will employ this factor to determine which news items
should be published prominently (Allern, 2002). Altogether, this guides us to the follow-
ing hypothesis:
H3: News factors (a) personification, (b) negativity, and (c) proximity will positively
relate to story prominence more strongly for popular press compared to quality press.
News factors and public versus commercial broadcasting
news
How public service broadcasters (PSB) and commercial broadcasters differ regarding the
value they assign to specific news factors is less clear. For several reason, news of the
PSB may follow a similar logic as quality news outlets – whereas commercial newscasts
may share similarities with popular outlets. PSBs have to meet the objectives formalized
in EU and national regulations (Hargreaves Heap, 2005): among others, aiding informed
citizenship through impartial and independent news by offering a wide range of quality
Boukes et al. 7
programs. Moreover, PSBs rely strongly on government funding and, hence, can be less
obsessed about making revenue through the inclusion of tabloid news factors. Thus simi-
lar to quality outlets, journalists working for the PSB’s newscast share a relative freedom
to be less profit-oriented, which makes a difference with commercial newscasters plau-
sible regarding the value assigned to specific news factors.
However, the quality criteria of governments for PSBs are often not clearly defined
(Hargreaves Heap, 2005) and their target audience is broad: News of public broadcasters
should not only inform with high quality content, but it should also inform a large share
of the citizenry. Although for a different motivation – public service versus selling adver-
tisements at a higher price – PSBs and commercial outlets, thus, partly share the same
objective. Moreover, PSBs are only partly financed through government funding in
many countries, with alternative income mostly coming from advertising. Such mixed
financing models open the door to a commercial logic (De Bens and Paulussen, 2005).
So, the PSB may also rely strongly on the news factors expected to be of particular
importance for popular news.
Furthermore, commercial broadcasters cannot differ too much from the public broad-
caster either, since they are often in competition with them for largely the same audience
(De Haan and Bardoel, 2012). Commercial news will therefore try to achieve similar
levels of quality coverage as the PSB (Nguyen and García-Martínez, 2012). In the coun-
try under investigation, the Netherlands, this also holds: RTL Nieuws, as the first com-
mercial broadcast news organization, is the largest competitor of the PSB’s flagship
news program NOS Journaal. Consequently, it competes for an audience that – for dec-
ades – has been socialized by the high standards of public broadcasting news. RTL
Nieuws, moreover, is subject to various systems of self-regulation and co-regulation. The
distinction between public television news bulletins and these of privately owned com-
mercial stations may, thus, not be as discernable as expected (Hendriks Vettehen et al.,
2005). In fact, previous research found that – given all the differences in terms of struc-
ture and organization – the coverage of public and commercial channels often is surpris-
ingly similar (De Bens and Paulussen, 2005; Kerbel et al., 2000). Thus, with theory
guiding us toward two opposite directions, we ask the following research question:
RQ2: Are specific news factors more or less strongly related to story prominence for the
newscasts of the public service broadcaster versus the commercial broadcaster?
Method
Data
Part of a larger research project (see, e.g., Boukes et al., 2019), a quantitative content
analysis was conducted on five complete months (1 February–8 July 2015) of economic
news coverage from 11 major Dutch news sources. Today, these news sources are still
among the most prominent and widely used ones. Four news outlets were identified as
popular news media: newspaper Algemeen Dagblad, newspaper Telegraaf and its online
counterpart telegraaf.nl, as well as the largest online news site with no offline equivalent
nu.nl.4 These popular news sources position themselves as ‘family’ newspapers
8 Journalism 00(0)
(Algemeen Dagblad) that ‘give people a voice’ (Telegraaf), with repeated emphasis on
the general character of their readership. Moreover, they generally contain briefer sto-
ries, larger headlines, and more colorful illustrations, which are all format characteristics
of tabloid journalism (Skovsgaard, 2014). In contrast, the two newspapers NRC
Handelsblad and Volkskrant, their respective online websites nrc.nl and volksrant.nl, as
well as the newspaper Trouw, describe themselves by their standards of ‘quality’
(Volkskrant) and ‘in-depth’ reporting (NRC and Trouw), aimed at ‘an audience that is
willing to think’ (Volkskrant). As such, these five news sources are defined as quality
news sources. Similar popular-versus-quality classifications of those newspapers were
made by Broersma and Graham (2012) as well as by d’Haenens and Bosman (2003).
Finally, a distinction was made between the PSB’s news bulletin NOS Journaal and the
privately owned newscast RTL Nieuws. For both, we use the prime-time evening news-
cast (respectively, 8PM and 7.30PM).
Units of analysis were the individual news items. All newspaper articles, all television
news items (n.b., not a sample) and 25 percent of all news website articles about the
economy were scrutinized for the presence of the seven news factors. Newspaper articles
were obtained with keyword searches on economic terms (see Boukes and Vliegenthart,
2017, or Supplemental Appendix A5) from the LexisNexis database and stored in the
Amsterdam Content Analysis Toolkit (Van Atteveldt et al., 2014). Economic news from
the websites were collected and stored within the Infrastructure for Automated Content
Analysis-tool (Trilling et al., 2018), after which the same search string was applied.
Irrelevant articles were removed by manually eliminating any article that in the headline
or first paragraph did not make explicit reference to economic issues. For the television
news broadcasts, all economic news items were manually identified (see instruction in
Supplemental Appendix A) through watching the full programs.
The total sample consisted of 4968 news articles among 11 Dutch national news
sources: 3691 newspaper articles, 207 television news items, and 1070 website articles.
Relatively limited overlap exists between print and online articles of newspapers. Similar
articles differ in length (Vandendaele, 2018), have different headlines, and partly differ-
ent content (see Boumans et al., 2018). On these grounds, we did not delete any similar
articles from the analysis. No major events occurred during our research period. The
most frequent topic of economic coverage was the Greek debt crisis: 16 percent of the
news items dealt with this issue. This is a substantial share of the news agenda, but still
a small minority of all items.
Measurements and reliability
Dependent variables. Two characteristics of a news item measured its prominence: length
and opening position. Opening position was defined as a dummy variable, with (1) for
articles placed on the front page of a newspaper or as the opening item in a newscast, and
(0) for those placed elsewhere. For websites, the homepage can be considered a struc-
tural equivalent of the front page. However, placement on websites is updated constantly,
more fluid, and harder to register; so, we had to exclude this aspect of the dependent vari-
able for the analysis of websites. Length was calculated as the number of words for a
news article or number of minutes for a TV item.
Boukes et al. 9
Independent variables. To measure the presence of news factors, a team of 22 student
coders conducted a manual content analysis. Table 1 contains a summary of the meas-
ured news factors, their definitions, general presence, and reliability statistics. Average
presence of the news factors is largely comparable with other studies analyzing the
presence of content features in news (e.g. De Keyser and Raeymaeckers, 2012; Harcup
and O’Neill, 2001). Intercoder reliability statistics are calculated with Nogrod 1.1
(Wettstein, 2018) available in the table, and further details are given in the Supplemen-
tal Appendix.
Analysis
Separate regression analyses were conducted for the three media types, because length is
conceptually different for textual (i.e. number of words) and audio-visual items (i.e.
number of minutes). Moreover, website articles were about 81.7 words shorter (p < .001)
than newspaper articles – complicating a one-on-one comparison. Length is a continuous
variable, thus, OLS regression models are used. Given the dichotomous nature of open-
ing position, logistic regression analyses were conducted predicting this dependent
variable.
The relationship between length and news factors is not unambiguous. As a proxy of
prominence, we argue that length is affected by news factors, but causality might be
reversed: Longer items have more space to contain more news factors. We cannot account
for this reversed causality directly, but given our focus on the relative importance of
news factors (and thus interaction terms), we can still assess whether the association
between each news factor and length differs across outlet types. For the other promi-
nence variable (opening position), length was included as control variable.
Results
News factors and prominence
A cumulative scale was generated indicating how many news factors were present in a
news item (M = 3.18, SD = 1.29, Min. = 0, Max. = 7). While controlling for outlet type,
logistic regressions showed that the odds of front-page publication increased with a fac-
tor 1.35 for every additional news factor present in an article (b = 0.30, SE = 0.09,
p = .001). The likelihood of being the opening item of a television newscast almost
doubled and increased with a factor 1.82 for every additional news factor present in a
story (b = 0.60, SE = 0.18, p = .001).
Length of newspaper articles increased with 92.15 words (SE = 3.86) with every addi-
tional news factor, and this relationship was strong (b* = .35, p < .001). A similar pattern
was found for the length of website articles (b = 68.85, SE = 7.64, b* = .25, p < .001). The
length of TV news items increased 0.73 minutes (SE = 0.10, b* = .44, p < .001) with every
additional news factor. Overall, these results confirm Hypothesis 1: More news factors
predict more upfront position and longer story length. Our method does not allow disen-
tangling the causal direction between the number of news factors and story length; we
reflect on this in the ‘Discussion’ section.
10 Journalism 00(0)
Table 1. News factors with their definitions, operationalizations, overall presence in economic news (% of news items), and intercoder-reliability
results.
News factor Definition Coding instruction Presence
(%)
Holsti (P.A.)
(%)
Std. Lotus
(λ)
Negativity Negative aspects of an
event or issue
Tone of article with regard to the economy coded as either
negative or mostly negative, mixed negative and positive (all
coded 1), neutral, mostly positive, positive, or no evaluation
(all coded 0)
11.7 79.3 0.75
Proximity Geographical nearness
of an event or issue
Story takes place in the Netherlands (1) or somewhere else
(0).
54.7 93.2 0.91
Eliteness Presence of
individuals,
organizations, or
nations of high status
involved in an event
or issue
Two items are combined indicating whether at least one of the
following actors was present (1):
a. Political elite: European Union, Dutch government, or
foreign government.
b. Economic elite: Dutch Bank, European Central Bank, labor
union, employers’ federation.
45.0 82.2 0.81
Influence and
relevance
High significance of
an event in terms
of its effects and/or
consequences
Measured by the presence of the ‘economic consequences-
frame’: Does the article explicitly refer to the economic impact
the story has, has had, or may have on individuals, companies,
a group, a religion, or a country?
81.2 75.3 0.73
Personification Inclusion of personal
or ‘human’ face of an
event/issue
Measured by two items:
a. Does the story use personal example(s) of an issue or
problem to illustrate trends in the economy?
b. Does the story feature a lay person or ‘man-on-the-street’?
10.3 86.7 0.89
Conflict Presence of
confrontation and/or
controversy
a. Does the story include disagreement between parties,
individuals, groups, organizations, or nations?
b. Does the story include an event or issue from two or more
sides?
46.2 78.4 0.66
Continuity Having been in the
news previously
The main topic was coded out of a list of 58 potential topics
(Std-λ = 0.79). Continuity was then operationalized as a news
item about a main topic that had also been covered in that
specific outlet in the 7 days before this publication.
67.3 n/a n/a
Boukes et al. 11
Which news factors matter?
To answer RQ1, we assessed how specific news factors related to the prominence of
news items. Table 2 shows the results for the two indicators of prominence (length and
front page/opening item) regarding the specific modalities. Overall, conflict had the
largest number of significant relationships: It related positively to the length of all
three the modalities, and increased the odds of a front-page publication in newspapers
with a factor 2.74. Similarly, eliteness increased the odds of front-page publication
with factor 2.10 and related positively to the length of textual news (print and online),
but not television news.
Overall, conflict and eliteness had the most unequivocal positive relationships with
the prominence indicators. Personification also related positively to story length in
newspapers, on websites, and in television news – but it negatively predicted the likeli-
hood of front-page publication; decreasing the odds with a factor 0.10. This raises the
question whether personification requires greater space (a practical consideration) rather
than that it boosts story newsworthiness. We elaborate on this in the ‘Discussion’ section.
Geographical proximity had a mixed impact. Although, it negatively related to the length
of website items, it related positively to story length in newspapers. Most interestingly,
proximity was the only news factor positively predicting the likelihood of being the
opening of a newscast (although marginally; p = .060).
The influence of three news factors was rather minimal: Influence and relevance,
negativity, and continuity. None of them significantly predicted the likelihood of upfront
position. Whereas, Influence and relevance still related significantly to the length of both
types of written news, negativity was only associated with the length of newspaper arti-
cles and continuity with length of TV news items.
The differential impact of news factors per outlet type
Interaction terms for every news factor are added to the regression models presented
earlier to examine whether relationships between specific news factors and prominence
are stronger for one type of outlet than the other.6 Outlet type is operationalized as a
dummy variable with value 0 for either the quality outlets or public broadcaster, and
value 1 for popular outlets or commercial broadcaster. Accordingly, the interaction effect
indicates whether a news factor’s main effect is stronger or weaker for popular/commer-
cial outlets vis-à-vis the quality/public outlets (i.e. the reference category). Table 3 shows
the results. Main effects of news factors and outlet type are not displayed for reasons of
clarity and space (i.e. main effects only indicate a news factor’s effect for the quality/
public outlets; the reference category), but are included in the model.
In most cases, no significant interaction effects were yielded. This means that, overall,
the different types of outlets were not significantly more or less sensitive to the presence
of the specific news factors. Nevertheless, a number of interesting significant interaction
relationships were revealed – in particular for the comparison of quality versus popular
newspapers. For example, regarding influence and relevance and story length: The nega-
tive interaction effect implies that the relationship is weaker for the popular newspapers
(see Figure 1). Story length increased with a 100 words for the quality outlet (from 427
12 Journalism 00(0)
to 527 words), whereas the increase in 30 words for the popular newspapers was not
significant. This confirms H2b. Conflict showed a similar pattern (i.e. negative interac-
tion effect): Its presence particularly increased story length in quality newspapers vis-à-
vis popular newspapers. Thus, confirming H2a.
A surprising finding was yielded for personification. As the only news factor that dif-
ferentially related to the story length on boht news websites and newspapers, its negative
interaction effects implied that personification particularly increased story length within
the quality outlets. Concretely, the presence of personification has an effect of 440 words
on length for the quality websites, but only increases article length with 52 words for the
popular outlets. Altogether, this was opposite to our expectation: H3b rejected.
Table 2. OLS and logistic regression models predicting news item prominence.
News factor Story length Upfront position
Presence Newspaper Website Television Newspaper Television
(0 = no, 1 = yes) b (SE)b (SE)b (SE) OR (SE) OR (SE)
Intercept 294.69*** 228.24*** 2.19** 0.01*** 0.00***
(15.78) (33.57) (0.65) (0.00) (0.00)
Conflict 141.58*** 108.22*** 0.80** 2.74*** 0.40
(10.53) (21.27) (0.26) (0.72) (0.26)
Eliteness 101.09*** 72.12*** 0.40 2.10** 1.99
(10.67) (21.42) (0.29) (0.54) (1.36)
Personification 252.33*** 292.85*** 1.45*** 0.10*0.40
(15.78) (50.29) (0.24) (0.10) (0.24)
Geographical
proximity
21.19* −43.65*−0.35 1.35 3.79†
(10.18) (20.79) (0.29) (0.31) (2.68)
Influence and
relevance
63.22*** 67.46*−0.69 0.86 3.23
(12.12) (29.37) (0.61) (0.25) (5.61)
Negativity 69.20*** 8.89 0.02 0.65 3.01
(15.16) (30.95) (0.43) (0.25) (2.54)
Continuity 13.93 26.56 0.78** 1.31 0.44
(10.70) (20.27) (0.27) (0.33) (0.29)
Popular news −191.12*** −87.25*** 0.62*
(9.76) (19.61) (0.15)
Commercial
broadcaster
−0.68*2.70†
(0.23) (1.55)
Length 1.00
(0.00)
4.37***
(1.10)
adj. R20.28 0.13 0.32
Log likelihood −24,977.20 −7387.40 −379.06 −387.75 −50.32
N3534 1035 202 3548 202
Cells contain unstandardized regression coefficients (b) or odds ratio (OR) with standard errors (SE) in
parentheses, and probabilities (p; two-tailed).
†p < .100; *p < .050; **p < .010; ***p < .001.
Boukes et al. 13
Although marginally significant, geographical proximity increased story prominence
of popular outlets both in terms of story length and upfront position (H4c tentatively
confirmed). While proximity increased story length with 38 more words for the popular
newspaper, it (marginally) increased the odds of a front-page publication in popular out-
lets with factor 2.68. Concretely, the likelihood was stable at 1.8 percent for the quality
outlets irrespective of whether proximity was presented, but the likelihood of front-page
publication increased from 0.8 percent to 2.1 percent for the popular newspapers when
geographical proximity was present.
Marginally significant interaction results were yielded for continuity and story length.
While it relatively increased story length for popular but not for the quality outlets (H2d
rejected), a positive relationship with length of TV news items only remained for the
public broadcaster. Altogether, the effect of most news factors, however, did not signifi-
cantly differ across the outlet types. Whereas still five differences were found between
Table 3. OLS and logistic regression models predicting prominence for different outlet types
(quality vs popular, or public vs commercial).
Intercept Story length Upfront position
Newspaper Website Television Newspaper Television
b (SE)b (SE)b (SE) OR (SE) OR (SE)
254.9*** 197.80*** 2.07*0.01*** 0.00*
(21.18) (47.88) (1.02) (0.01) (0.01)
Conflict ×
popular/commercial
−70.77*** −37.81 −0.38 0.48 0.91
(20.93) (42.43) (0.53) (0.26) (1.19)
Eliteness ×
popular/commercial
−4.81 16.41 −0.38 0.80 4.25
(21.23) (42.91) (0.59) (0.42) (6.28)
Personification × popular/
commercial
−171.05*** −387.50*** −0.51 – 4.10
(31.40) (102.9) (0.47) – (5.29)
Geographical proximity ×
popular/commercial
38.17†54.32 −0.00 2.68†6.04
(20.27) (41.84) (0.59) (1.37) (9.23)
Influence and relevance ×
popular/commercial
−69.81** −39.33 0.70 0.67 −
(24.12) (58.86) (1.28) (0.40) −
Negativity ×
popular/commercial
−17.22 −55.39 0.01 0.77 0.10
(30.11) (61.91) (0.95) (0.61) (0.22)
Continuity ×
popular/commercial
35.69†−15.41 −0.92†0.99 3.23
(21.39) (40.45) (0.55) (0.53) (4.38)
Length 1.00
(0.00)
5.98***
(1.91)
adj. R20.29 0.14 0.33
Log likelihood −24,943.81 −7377.41 −374.53 −383.00 −45.04
N3534 1035 202 3371 202
Cells contain unstandardized regression coefficients (b) or odds ratio (OR) with standard errors (SE) in
parentheses, and probabilities (p; two-tailed). Main effects are not displayed, but included in model.
†p < .100; *p < .050; **p < .010; ***p < .001.
14 Journalism 00(0)
the popular versus quality newspapers, the websites only differed with respect to the
impact of personification, and newscasts only differed regarding continuity. This means
that none of the hypotheses regarding the diffential impact of specific news factors
received full support; only partial evidence was found for influence and relevance, con-
flict, and proximity in printed newspapers.
Discussion
This study illustrates that the presence of a greater number of news factors was unequivo-
cally related to both story length and upfront position across three modalities (newspaper,
websites, television). This reveals the potential leverage news factors may have on news-
worthiness and, thus, on journalistic decisions about prominence in particular. To our
knowledge, this is the first time that the causal mechanisms predicted by news value theory
(Galtung and Ruge, 1965; Staab, 1990) have been exposed on such a large scale, although
still regarding prominence instead of the selection of news. Theoretically, our study puts
empirical flesh on the bones of news value theory by moving beyond the mere presence of
news factors and demonstrating their potential consequences for story prominence.
Our study, moreover, demonstrates that not all news factors contribute equally to the
prominence of stories. Increases in prominence were mostly found for news items that
contained conflict and eliteness: They both positively predicted the likelihood of front-
page publication and story length. Proximity had a mixed effect: While it increased story
length in offline newspapers, it decreased length in online newspapers. Apparently,
online outlets seem to only cover short (factual) foreign news, whereas printed newspa-
pers may be more prone to pick up (background) stories from other international media.
Proximity was also the only news factor increasing the odds of being the opening of a
newscast.
Figure 1. Predicted length as a function of outlet type and presence of news factor “influence
and relevance” in newspaper articles.
Boukes et al. 15
Three news factors (i.e. influence and relevance, negativity, continuity) did not
increase the odds of a story’s upfront position, but only related to story length of one or
two news modalities. Their influence, thus, seems rather limited. Notably, personifica-
tion related positively to story length for newspapers, websites, and television – but
negatively predicted the likelihood of front-page publication. Arguably, this news factor
may increase the length of news because longer stories are simply required to include the
perspectives of the general public, and not necessarily because they are deemed more
newsworthy.
Similarly, quotes of elites or the interpretation of conflict require space, whereas to
frame something as negative or the choice for a (domestic) topic that had been in the
news before (i.e. continuity) does not necessarily involve additional words. The causal
direction between the presence of specific news factors and story length, thus, could not
be fully disentangled in our content analysis. In contrast, reverse causality is less likely
for the relationship between the presence of specific news factors and the odds of upfront
position, especially because the analyses controlled for story length. Conflict and elite-
ness, therefore, seem to be the most influential news factors; these affected both length
and position.
The literature made us expect differences between the influences of news factors by
outlet type. However, our findings did not strongly support this expectation: By contrast,
most interaction effects revealed no conditionality upon outlet type. The lack of signifi-
cant interaction effects can be explained journalistic organizations’ tendency to ‘de-dif-
ferentiate’ and to follow the dominant logic of the field to seek or maintain legitimation
as a news source (Tandoc, 2018). Another explanation could be the similar kinds of
training that journalists of the different outlets may receive (Parks, 2019). Hence, the
objective of reader appeal is not only the ultimate production value of commercial out-
lets, but also for quality broadsheets (see Vandendaele, 2018), and the same news factors
may be used to achieve this.
One could question the generalizability of these results outside the Netherlands.
Greater differences between commercial and public broadcasting news may emerge in
countries where the PBS is fully government funded (e.g. the Scandinavian countries)
and does not rely on advertising revenues as the Dutch PBS does (Saurwein et al., 2019):
Only 80 percent of its budget is government funded and government spending on public
broadcasting is among the lowest across Europe. Moreover, it is conceivable that greater
differences exist in the susceptibility to news factors in media landscapes with less regu-
lations (the US), with more extreme forms of tabloid journalism (the UK and Germany)
or with public broadcasters that have lower market shares (Portugal or Greece, see
Picard, 2002).
Hardly any differences emerged between the popular and quality news websites in our
study. A possible explanation could be the universally strong reliance on pre-produced
news of wire services in online news (Boumans et al., 2018), which leaves little agency
to the journalists in the ‘online’ newsroom. Another question is of how the discursive
construction of economic newsworthiness may differ from other news topics. It might
well be that coverage of other topics is more/less sensitive to certain news factors. One
can imagine that a factor such as personification is stronger and more influential if the
16 Journalism 00(0)
focus is on ‘softer’ issues, whereas eliteness could be more influential when focusing on
‘harder’ issues.
Future research should gather insights from within the newsroom to add a production
perspective to our product-only perspective (Catenaccio et al., 2011). A linguistic ethno-
graphic approach (e.g. Van Hout and Macgilchrist, 2010) that examines journalistic work
within its institutional context could verify the existence of and disentangle causality in
the relationship between news factors and prominence that we revealed. Our large-scale
analysis, necessarily, oversimplified the complex processes through which news is pro-
duced and largely ignores the (interactive) dynamic role of journalism professionals
(Catenaccio et al., 2011). Moreover, our binary measurement of some of the news factors
– e.g., negativity and proximity (that basically captures the difference between domestic
and foreign news) – could be more fine-grained in future research. This is even more true
for conflict, where we used a very broad conceptualization that might be in line with
previous research, but does not do right to the multifaceted nature of the phenomenon
(Feldman and Warfield, 2010).
Altogether, our study is of theoretical significance by empirically examining
whether, how and which news factors affect the journalistic process of determining the
prominence that should be given to news stories. Thereby, we have examined but could
not confirm previous notions about the contextual influence of news factors; only few
differences emerged between different outlet types – mainly between the quality ver-
sus popular newspapers, but hardly any differences were found between websites or
public versus commercial broadcasters. Our insights would benefit from verification
by newsroom insiders and the producers of these news artifacts to gain a more thor-
ough understanding of the differential processes that translate news factors into media
content.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/
or publication of this article: This work was supported by the Netherlands Organization for
Scientific Research (NWO) with a VIDI grant under project number: 016.145.369. Data collection
of the online news was carried out via INCA on the Dutch national e-infrastructure with support of
the SURF Cooperative.
ORCID iD
Mark Boukes https://orcid.org/0000-0002-3377-6281
Supplemental material
Supplemental material for this article is available online.
Notes
1. Although the terms ‘popular’ and ‘quality’ have their ambiguities, and are often interchange-
ably used with the terms ‘tabloid’ versus ‘broadsheet’ newspapers, we prefer the first, because
the latter distinction may be confused with the physical size of outlets. Moreover, none of
the Dutch popular newspapers is comparable to ‘real’ tabloid outlets, such as the Sun or Bild.
Boukes et al. 17
With popular press, we refer to newspapers that are supposed to ‘address issues of direct
concern to ordinary people, to display an anti-elite bias and to focus on conflict and drama’
(Akkerman, 2011: 931), while quality newspapers are argued to provide more comprehen-
sive, in-depth coverage, and employ official frames that privilege the political elites’ interpre-
tations of issues or events (Akkerman, 2011: 934).
2. Additional analyses show that a second (societal) issue next to the main economic issue is
more often present in quality press (49.4%) than in popular press (36.1%).
3. Additional analyses show that the news factors of ‘eliteness’ and ‘influence and relevance’
also correlate with each other: r* = .39 (i.e. tetrachoric correlation for binary variables).
4. Practical difficulties (cookie wall, irregular lay-out) hampered scraping AD.nl, Algemeen
Dagblad’s website.
5. Supplemental Appendix available at https://osf.io/y54wu/?view_only = fbad68fc5c2b410
9b5d1d27e9095f554
6. The interaction effect of personification and outlet type for newspapers on front page could
not be examined, because none of the articles on the front page of popular newspapers (n = 32)
contained personification. Similarly, it was impossible to examine the interaction effect of
‘influence and relevance’ by outlet type for television news because every commercial broad-
cast opening item contained influence and relevance (n = 19).
References
Akkerman T (2011) Friend or foe? Right-wing populism and the popular press in Britain and the
Netherlands. Journalism 12(8): 931–945.
Allern S (2002) Journalistic and commercial news values. Nordicom Review 23(1–2): 137–152.
Barnett S and Seymour E (1999) A Shrinking Iceberg Travelling South: Changing Trends in
British Television. London: Campaign for Quality Television.
Bartholomé G, Lecheler S and de Vreese C (2015) Manufacturing conflict? How journalists inter-
vene in the conflict frame building process. The International Journal of Press/Politics 20(4):
438–457.
Bednarek M and Caple H (2014) Why do news values matter? Discourse & Society 25(2): 135–158.
Boukes M and Vliegenthart R (2017) A general pattern in the construction of economic news-
worthiness? Analyzing news factors in popular, quality, regional, and financial newspapers.
Journalism. Epub ahead of print 31 August. DOI: 10.1177/1464884917725989.
Boukes M, Damstra A and Vliegenthart R (2019) Media effects across time and subject: How
news coverage affects two out of four attributes of consumer confidence. Communication
Research. Epub ahead of print 24 August 2019. DOI: 10.1177/0093650219870087.
Boumans J, Trilling D, Vliegenthart R, et al. (2018) The agency makes the (online) news world go
round: The impact of news agency content on print and online news. International Journal of
Communication 12: 1768–1789.
Broersma M and Graham T (2012) Social media as beat. Journalism Practice 6(3): 403–419.
Caple H and Bednarek M (2013) Delving into the Discourse: Approaches to News Values in
Journalism Studies and beyond. Oxford: Reuters Institute for the Study of Journalism.
Catenaccio P, Cotter C, De Smedt M, et al. (2011) Towards a linguistics of news production.
Journal of Pragmatics 43(7): 1843–1852.
Cotter C (2010) News Talk: Investigating the Language of Journalism. Cambridge: Cambridge
University Press.
Curran J, Douglas A and Whannel G (1980) The political economy of the human interest story.
In: Smith A (ed.) Newspapers and Democracy: International Essays on a Changing Medium.
Cambridge, MA: The MIT Press, pp. 288–347.
18 Journalism 00(0)
d’Haenens L and Bosman J (2003) Portrayal of Canada in the Dutch print media. Canadian
Journal of Communication 28(2): 203–218.
De Bens E and Paulussen S (2005) Hoe anders is de VRT? De performantie van de vlaamse pub-
lieke omroep. Tijdschrift Voor Communicatiewetenschap 33(4): 365–386.
De Haan Y and Bardoel J (2012) Accountability in the newsroom: Reaching out to the public or a
form of window dressing? Studies in Communication Sciences 12(1): 17–21.
De Keyser J and Raeymaeckers K (2012) The printed rise of the common man. Journalism Studies
13(5–6): 825–835.
Eilders C (2006) News factors and news decisions: Theoretical and methodological advances in
Germany. Communications 31(1): 5–24.
Elorza I (2014) Newsworthiness, attribution and lexicogrammatical strategies in two types of news
articles in English and Spanish. Topics in Linguistics 14(1): 16–33.
Feldman R and Warfield TA (2010) Disagreement. New York: Oxford University Press.
Fico F and Freedman E (2001) Setting the news story agenda: Candidates and commentators in
news coverage of a governor’s race. Journalism & Mass Communication Quarterly 78(3):
437–449.
Galtung J and Ruge MH (1965) The structure of foreign news. Journal of Peace Research 2(1):
64–91.
Harcup T and O’Neill D (2001) What is news? Galtung and Ruge revisited. Journalism Studies
2(2): 261–280.
Harcup T and O’Neill D (2017) What is news? News values revisited (again). Journalism Studies
18(12): 1470–1488.
Hargreaves Heap SP (2005) Television in a digital age: What role for public service broadcasting?
Economic Policy 20(41): 112–157.
Hendriks Vettehen P, Nuijten K and Beentjes J (2005) News in an age of competition: The case of
sensationalism in Dutch television news, 1995–2001. Journal of Broadcasting & Electronic
Media 49(3): 282–295.
Jacobs G and Tobback E (2013) Is language a news value in Belgium? Journalism Studies 14(3):
407–422.
Kepplinger HM and Ehmig SC (2006) Predicting news decisions. An empirical test of the two-
component theory of news selection. Communications 31(1): 25–43.
Kerbel MR, Apee S and Ross MH (2000) PBS ain’t so different: Public broadcasting, election frames,
and democratic empowerment. Harvard International Journal of Press/Politics 5(4): 8–32.
Lee NY and Baek K (2018) Squeezing out economic news for business news? Journalism 19(9–
10): 1220–1238.
Masterton M (2005) Asian journalists seek values worth preserving. Asia Pacific Media Educator
16: 41–48.
Mitchell A and Holcomb J (2016) State of the News Media 2016. Washington, DC: Pew Research
Center.
Mothes C, Knobloch-Westerwick S and Pearson GD (2019) The PFAD-HEC model: Impacts of
news attributes and use motivations on selective news exposure. Communication Theory
29(3): 339–359.
Nguyen A and García-Martínez AN (2012) When public service is the name of the game. In: Burns
M and Brügger N (eds) Histories of Public Service Broadcasters on the Web. New York:
Peter Lang, pp. 1–16.
Parks P (2019) Textbook news values: Stable concepts, changing choices. Journalism & Mass
Communication Quarterly 96(3): 784–810.
Picard RG (2002) Assessing audience performance of public service broadcasters. European
Journal of Communication 17(2): 227–235.
Boukes et al. 19
Reinemann C, Stanyer J, Scherr S, et al. (2012) Hard and soft news: A review of concepts, opera-
tionalizations and key findings. Journalism 13(2): 221–239.
Saurwein F, Eberwein T and Karmasin M (2019) Public service media in Europe: Exploring
the relationship between funding and audience performance. Javnost – The Public 26(3):
291–308.
Schulz WF (1982) News structure and people’s awareness of political events. Gazette 30(3):
139–153.
Semetko HA and Valkenburg PM (2000) Framing European politics: A content analysis of press
and television news. Journal of Communication 50(2): 93–109.
Skovsgaard M (2014) A tabloid mind? Professional values and organizational pressures as expla-
nations of tabloid journalism. Media, Culture & Society 36(2): 200–218.
Staab JF (1990) The role of news factors in news selection: A theoretical reconsideration. European
Journal of Communication 5(4): 423–443.
Strömbäck J, Karlsson M and Hopmann DN (2012) Determinants of news content. Journalism
Studies 13(5–6): 718–728.
Tandoc EC (2018) Five ways BuzzFeed is preserving (or transforming) the journalistic field.
Journalism 19(2): 200–216.
Tresch A (2009) Politicians in the media: Determinants of legislators’ presence and prominence in
Swiss newspapers. International Journal of Press/politics 14(1): 67–90.
Trilling D, Tolochko P and Burscher B (2017) From newsworthiness to shareworthiness: How
to predict news sharing based on article characteristics. Journalism & Mass Communication
Quarterly 94(1): 38–60.
Trilling D, Van de Velde B, Kroon AC, et al. (2018) INCA: Infrastructure for content analy-
sis. In: 2018 IEEE 14th international conference on e-science, Amsterdam, 29 October–1
November, pp. 329–330. New York: IEEE.
Tumber H (1993) ‘Selling scandal’: Business and the media. Media, Culture & Society 15(3):
345–361.
Van Atteveldt W, Ruigrok N, Takens J, et al. (2014) Inhoudsanalyse Met Amcat. Available at:
http://vanatteveldt.com/wp-content/uploads/amcatbook.pdf
Van Hout T and Macgilchrist F (2010) Framing the news: An ethnographic view of business news-
writing. Text & Talk 30(2): 169–191.
Vandendaele A (2018) ‘Trust me, I’m a sub-editor’. Journalism Practice 12(3): 268–289.
Wettstein M (2018) Nogrod 1.1 (beta): Quick tutorial. Available at: https://www.ikmz.uzh.ch/dam/
jcr:fd26fbd3-2b3f-4421-96f8-4c2750417551/Nogrod_1-1.pdf (accessed 8 January 2019).
Author biographies
Mark Boukes is an assistant professor in the department of Communication Science at the
Amsterdam School of Communication Research (ASCoR), University of Amsterdam. His research
focuses on media content and effects of journalistic versus infotainment. Moreover, Mark investi-
gates the content and consequences of economic news coverage.
Natalie P Jones is a Research Master student in the Graduate School of Communication, University
of Amsterdam. She focuses on political communication and journalism, with a special focus on
climate and immigration.
Rens Vliegenthart is a full professor for Media and Society in the department of Communication
Science and at the Amsterdam School of Communication Research (ASCoR), University of
Amsterdam (UvA). His research focuses on media–politics relations, media coverage of social
movements and businesses, election campaigns, and economic news coverage.
