ArticlePDF Available

Abstract and Figures

Internet users are constantly confronted with metric information about the popularity of goods, services, and content. These popularity cues (PCs) - which we define as metric information about users' behavior or their evaluations of entities - serve as social signals for users who are confronted with them. Due to the high relevance that PCs have for organizations, consumers, and scholars, this article provides a systematic overview of PC research. First, we present a theoretical conceptualization for the effects of PCs. Second , we analyze empirical research that focuses on PCs by providing a review of academic, peer-reviewed studies on the direct effects of PCs in online media (N = 61). Third, we utilize the results of our literature review to address current shortcomings in the literature and to provide insights for future research.
Content may be subject to copyright.
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018, S. 186–207, DOI: 10.5771/2192-4007-2018-2-186
SC
|
M
Studies in Communication and Media
FULL PAPER
Popularity cues in online media:
A review of conceptualizations, operationalizations,
and general effects
Popularitätshinweise in Online-Medien:
Ein systematischer Überblick über Konzeptualisierungen,
Operationalisierungen und Effekte
Mario Haim, Anna Sophie Kümpel & Hans-Bernd Brosius
SC
|
M
Studies in Communication and Media
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

SC
|
M
Studies in Communication and MediaStudies in Communication and Media
Mario Haim (M.A.), Institut für Kommunikationswissenschaft und Medienforschung,
Ludwig-Maximilians-Universität München, Oettingenstr. 67, 80538 München, Germany;
Contact: haim(at)iw.lmu.de
Anna Sophie Kümpel (M.A.), Institut für Kommunikationswissenschaft und Medien-
forschung, Ludwig-Maximilians-Universität München, Oettingenstr. 67, 80538 München,
Germany; Contact: anna.kuempel(at)iw.lmu.de
Hans-Bernd Brosius (Prof. Dr.), Institut für Kommunikationswissenschaft und Medien-
forschung, Ludwig-Maximilians-Universität München, Oettingenstr. 67, 80538 München,
Germany; Contact: brosius(at)iw.lmu.de
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
FULL PAPER

FULL PAPER
Popularity cues in online media: A review of conceptualizations,
operationalizations, and general effects
Popularitätshinweise in Online-Medien: Ein systematischer
Überblick über Konzeptualisierungen, Operationalisierungen und
Effekte
Mario Haim, Anna Sophie Kümpel & Hans-Bernd Brosius
Abstract: Internet users are constantly confronted with metric information about the pop-
ularity of goods, services, and content. These popularity cues (PCs)—which we dene as
metric information about users’ behavior or their evaluations of entities—serve as social
signals for users who are confronted with them. Due to the high relevance that PCs have
for organizations, consumers, and scholars, this article provides a systematic overview of
PC research. First, we present a theoretical conceptualization for the effects of PCs. Sec-
ond, we analyze empirical research that focuses on PCs by providing a review of academic,
peer-reviewed studies on the direct effects of PCs in online media (N = 61). Third, we uti-
lize the results of our literature review to address current shortcomings in the literature
and to provide insights for future research.
Keywords: Literature review, popularity cues, online media, social media
Zusammenfassung: Internetnutzer_innen werden fortlaufend mit aggregierten Daten über
die Beliebtheit von Gütern, Dienstleistungen oder (Medien-)Inhalten konfrontiert. Diese
Popularitätshinweise (PH), die wir als metrische Informationen über das Verhalten von
Nutzer_innen oder deren Bewertung von Entitäten denieren, fungieren als (soziale) Signa-
le, an denen sich Anwender_innen orientieren können. Angesichts der hohen Relevanz von
PH für Organisationen, Konsument_innen und nicht zuletzt Forscher_innen bietet dieser
Beitrag einen Überblick über die Forschung zu PH. Wir stellen dafür 1) Überlegungen zu
einer theoretischen Verankerung von PH an, geben 2) mithilfe einer systematischen Litera-
tursynopse bestehender Studien (N=61) einen Einblick in aktuelle Forschungsarbeiten
und nutzen 3) die Befunde unseres Reviews, um bestehende Probleme in der PH-Forschung
zu adressieren und Empfehlungen für künftige Forschungsvorhaben zu formulieren.
Schlüsselwörter: Literatursynopse, Popularitätshinweise, Online-Medien, Soziale Medien
Note: This research was supported by the Munich Center for Internet Research
(MCIR) and the German Research Foundation [research group “Political Com-
munication in the Online World,” subproject 1, grant number 1381]
Popularity cues in online media – a review
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
1. Introduction
Internet users are ubiquitously provided with metric information about the popu-
larity of online goods, services, and content, such as 900,000 Likes for Mark
Zuckerberg’s latest Facebook post, an average of 2 out of 5 stars by 75 users for
a restaurant on Yelp, or a 9.1 lm rating by 23,000 movie fans on IMDB. Users
can also provide metric ratings and thus contribute to the bulk of available user
experiences with products and services. These popularity cues (PCs) directly serve
as social signals for users who are confronted with them (Ksiazek, Peer, & Les-
sard, 2016). Moreover, such PCs are put to algorithmic use within ltered online
environments such as social network sites; thus, they can also serve as indirect
signals for users through the selection and arrangement of information based on
its popularity (Napoli, 2010). As prior research shows, PCs are both directly and
indirectly able to inuence users’ perceptions of the entity associated with PCs.
Hence, they might affect the users’ subsequent decisions in terms of selection, us-
age, and evaluations.
In research, however, PCs suffer from strong conceptual and operational ambi-
guity. As such, they are operationalized as both independent and dependent varia-
bles. While, strictly speaking, the term ‘cue’ is misleading due to its connotation of
decisiveness, which suggests an effect on users, a plethora of terms are used to de-
note PCs as independent variables. The variety of terms ranges from “bandwagon
cues” (Kim & Sundar, 2014) and “helpfulness ratings” (Walther, Liang, Ganster,
Wohn, & Emington, 2012) to “social media metrics” (Stavrositu & Kim, 2014) and
“social endorsement cues” (Messing & Westwood, 2014, p. 1046). Theoretical con-
ceptualizations of PCs also vary, including arguments related to word-of-mouth
(Duan, Gu, & Whinston, 2008), involvement (Kim, Brubaker, & Seo, 2015), exem-
plication (Peter, Rossmann, & Keyling, 2014), and news-value (Weber, 2014)
theory. Moreover, PCs are presented in either real-world (e.g., Facebook) or cti-
tious environments, are visualized either graphically (e.g., star ratings) or numeri-
cally, refer to actual (e.g., Likes for a post) or follow-up (e.g., Likes for a comment)
content, describe content (e.g., Likes for a post), usage (e.g., number of clicks), dif
-
fusion (e.g., number of shares), or follow-up communication (e.g., number of com-
ments), and possess either an evaluative character (e.g., Likes) or are non-evaluative
per se (e.g., clicks). However, a systematic overview of the conceptualizations,
operationalizations, and general effects of PCs is still missing.
A systematic overview of PCs is crucial for a coherent understanding of how
the perception of others’ behaviors and evaluations could affect individual users
under various circumstances. Many studies refer to PCs as a central feature of
social network sites, news aggregation, and e-commerce. Yet, this strong depend-
ence on context results in the fragmentation of conceptual assumptions, thus
hampering a comprehensive perspective. In those studies, the equivocal variety of
conceptualizations and operationalizations allows for cherry-picking of suitable
ndings. It does not facilitate a systematic overview of the possible effects of PCs.
For example, while some empirical ndings have shown that PCs affect users’
news selection (e.g., Yang, 2016), other studies suggest that PCs only have a very
limited effect (e.g., Knobloch-Westerwick, Sharma, Hansen, & Alter, 2005).
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
Moreover, various moderating inuences have been identied, but these have only
been discussed from the perspective of highly specic scenarios.
Thus, this paper aims to provide a generalizable overview of the effects of PCs
in online media that were studied up to this point. To do so, we rst embed the
concept of PCs within the broader contexts of relevance cues and attentional pro-
cesses. Building upon this theoretical foundation, we provide a review of aca-
demic, peer-reviewed studies to systematically collect and analyze empirical nd-
ings on the direct effects of PCs in online media (N= 61). Coding included all
aspects of the empirical studies that are both comparable and relevant from a
media effects perspective. This included the main eld of interest, the methodol-
ogy, operationalization, dependent and moderating variables, and outcome. How-
ever, due to the wildly varying methodological approaches used in the reviewed
studies, we could not apply common meta-analytic procedures. Rather, we pro-
vide a quantitative descriptive overview of the investigated studies. Ultimately, we
utilize the results of our review to address current shortcomings in the literature
and to provide insights for future research.
1.1 Popularity cues in online media
Despite the variety of terms and theoretical conceptualizations, scholars agree on
various principles with regard to PCs. First, PCs represent meta-information
about the popularity of an entity (e.g., a product, social-media post, or news arti-
cle). By itself, meta-information is neither inherent to nor entirely dependent on
an entity’s manifest characteristics. From a general perspective, the informational
value of PCs merely can be seen as a cue for further interpretation (for a literature
overview on PCs as results of prior behavior, see Porten-Cheé et al.’s paper in this
issue). For example, a news article might get several thousand Likes on Facebook,
whereas the exact same article might only receive a few Hearts on Twitter. Sec-
ond, PCs reveal either intended user-generated information (e.g., ratings) or unin-
tended (observed) user-generated information (e.g., number of clicks). Yet, in real-
ity, PCs do not necessarily rely on or reveal this discrimination. Thus, their value
could be user-generated, observed, or a (nontransparent) combination of the two.
Third, PCs depict metrics rather than qualitative data (e.g., comments). That said,
PCs are not necessarily presented as plain numbers. Instead, they might also be
illustrated, for instance as an icon or as a graphic image.
In line with these concurring principles, we dene PCs as metric information
about users’ behavior or their evaluations of entities. However, the term popularity
requires further clarication. First, popularity implies an indication of relevance,
be it positive or negative, among a population. Second, it refers to a population
among which the perceived popularity is valid. This population can be known or
unspecic. Moreover, it could either be platform-driven (e.g., popular on Twitter)
or externally constructed (e.g., popular among U.S. citizens), both of which are
subject to individual interpretations. That being said, PCs do not per se indicate
the same kind and amount of relevance to every user. Thus, the meaning of PCs—
whether they are ‘high’ or ‘low’ or whether they indicate favorable or unfavorable
evaluations—can only be ascribed by users and their individual assessments. This
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
process includes (unconsciously) weighting PCs against each other, incorporating
prior knowledge, or considering one’s presumptions about the evaluated entity.
Therefore, PCs can be understood and categorized under the umbrella concept of
relevance cues, which, depending on the users’ individual assessments of the PCs
under consideration, may or may not affect their evaluations of a given entity.
However, the concept of relevance cues is neither necessarily limited to online or
metric information nor does it solely serve as an indication of popularity.
1.2 Popularity cues as relevance cues
Relevance cues are indicators that signal a certain level of importance to media
recipients. They offer information regardless of the actual elaboration of the con-
tent. Relatedly, peripheral cues refer to indicators that trigger heuristic content
elaboration, but these cues do not necessarily depict relevance (Petty & Caciop-
po, 1986). Apart from that, to the best of our knowledge, no systematic differen-
tiation of relevance cues exists. Thus, we distinguish between four types of rele-
vance cues. First, internal relevance cues designated by the originator of a message
include all kinds of signals that are intentionally included in an entity (e.g., in a
news article or a product description) to indicate importance, such as highlighted
news values (e.g., “the biggest environmental disaster in human history”), celeb-
rity endorsements (e.g., “Rihanna supports this campaign”), or linguistic features
(e.g., exclamation marks). Second, external relevance cues designated by the orig-
inator of a message include signals that are intentionally attached to an entity and
indicate importance relative to other entities. Such cues include labels (e.g., “edi-
tor’s pick”), layouts, or an item’s ranking on a website. Third, external relevance
cues designated by intermediaries depict intentionally attached signals to an enti-
ty by a third party that is neither the originator nor user of a message (Helmond,
2015; Nielsen & Ganter, 2017). Examples include algorithmically derived rank-
ings or personalization features which present information because they suppos-
edly t users’ preferences. Fourth, external relevance cues designated by users are
signals attached to an entity that are intentionally or unintentionally produced
and curated by recipients or consumers. In contrast to relevance cues designated
by either the originator of a message or intermediaries, relevance cues designated
by users indicate a level of popularity among those users. In the context of TV
talk shows, Nabi and Hendriks (2003) referred to these types of cues as audience-
response cues. While these might include live reactions on TV or radio, such as
applause, individual close-up reactions, or telephone polls, our more general un-
derstanding also includes online reactions, such as comments or metric informa-
tion about users’ behavior or their evaluations of entities—that is, PCs.
The possible effects of these types of relevance cues may differ. While internal
and external relevance cues designated by the originator of a message as well as
external relevance cues designated by intermediaries (i.e., types one, two, and
three) suggest that users follow the originator’s guidance, external relevance cues
designated by users indicate broader opinions that, in turn, might be perceived as
more independent and diverse. Because it is generally assumed that people surveil
their environment in order to perceive public opinion, people build their percep-
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
tions on cues that indicate relevance (Hardmeier, 2008; Noelle-Neumann, 1974).
External relevance cues designated by users thus have the potential to affect per-
ceptions of public opinion (see Porten-Cheé et al.’s related discussion in this is-
sue). Despite the potential effectiveness of relevance cues, individuals’ perceptions
of relevance cues are neither static nor immutable. Rather, they evolve over time,
depending on individual usage patterns, technological capabilities, and societal
assumptions. For example, most likely, the number of Likes will be associated
with ‘positive popularity’ (acclaim, approval), whereas the number of ‘angry
emoticons’ may be associated with ‘negative popularity’ (blame, disapproval). In
the remainder of this paper, we focus on the current conceptualizations of PCs as
external relevance cues designated by users.
1.3 Popularity cues and attentional processes
To the best of our knowledge, no dedicated theoretical conceptualizations of the
attentional processes associated with PCs exist within communication studies.
Yet, the eld of social cognition offers insights into people’s information process-
ing, which may help explain how PCs could affect a) attentional processes and,
subsequently, b) the formation of users’ impressions.
Inherently, the perception of information begins with attention toward said
information (Bodenhausen & Hugenberg, 2009). Due to limitations of cognitive
capacity, attention can only be ascribed selectively (Posner, 1994). Which infor-
mation receives (selective) attention is subject to a process that involves a broad
variety of inuences, and it starts with “preattentive scans of the environment”
(Bodenhausen & Hugenberg, 2009, p. 4). According to Bodenhausen and Hugen-
berg (2009), information either grabs a person’s attention (bottom-up) or a per-
son actively seeks certain information (top-down). Once an entity is within a per-
son’s subconscious attention, various (contradicting) evaluation mechanisms
come into play. In terms of media content and the formation of people’s impres-
sions, three concepts address such evaluation mechanisms: vividness, salience, and
differential attention.
First, vividness has served as a discriminating cue in which content is perceived
as either lively and worth remembering or dull and apt to be ignored (Kisielius &
Sternthal, 1984, 1986; Taylor & Thompson, 1982). In this regard, popularity is
directly attached to an entity. Thus, it represents an absolute measure because it
allows an entity to be rated as vivid without comparing it to another entity. We
call this a between-subjects indication of popularity (i.e., a bottom-up signal for
attention). In order for PCs to act in this way, a consensus would be necessary in
which the recipients, the originators, and the researchers agree on ‘high’ and ‘low’
levels of PCs. Yet, while this sometimes is possible (e.g., ve stars are generally
more captivating than three stars), oftentimes, and especially with raw numbers,
this is not the case because PCs depend on the perceived characteristics of an en-
tity. For instance, while 230 product reviews might be ‘a lot’ when considering
buying a new belt, it may as well be ‘not much’ when it comes to a new smart-
phone. Thus, vividness is an approach that cannot solely explain the attentional
processes prompted by PCs.
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
Second, the concept of salience describes the relevance people ascribe to issues
1
.
Among other factors, salience subsequently leads people to derive rank-orders;
thus, it is attached to an individual person rather than the issue itself (Evatt, 1997;
Kiousis, 2004). We refer to this as a within-subject indication of popularity (i.e., a
top-down attribution of attention). For example, while one person might consider
6 out of 10 points to be a high rating, someone else might nd a minimum value
of 8 points to be acceptable. Moreover, salience is likely to vary systematically
within subjects, depending on the PCs under consideration. A person might rely on
the rating for movies (e.g., 6 out of 10 points), whereas the same individual might
primarily focus on the amount of ratings for printer supplies (e.g., 230 ratings).
While salience seems widely applicable to the concept of PC-driven attentional
processes, it ignores the inuence of content-specic characteristics.
Third, for the analysis of attentional processes to media stimuli, Brosius and
Mundorf (1990; original publication in German) describe a concept they call dif-
ferential attention2. They suggested looking at both vividness and salience simul-
taneously when analyzing attentional processes to media stimuli, because in real
life neither of the concepts occurs in isolation. However, in the past, researchers
have primarily examined vividness and salience in separate studies, which increas-
es the risk of confounding. For example, in experimental vividness studies, two
groups of participants are often presented with vividly diverging stimuli, but the
differences in the studies’ outcomes may also be due to variations in individual
salience. Following ideas from the eld of social cognition (Nisbett & Ross,
1980), Brosius and Mundorf (1990) noted that it is important to understand the
use of media content as an integrated combination of content-specic aspects
(i.e., between-subjects vividness) and cognitive aspects (i.e., within-subject sali-
ence). Moreover, Brosius and Mundorf (1990) suggested including culture-bound
aspects, which were already proposed as inuential aspects within the news-value
theory (Galtung & Ruge, 1965). For instance, to a movie enthusiast (salience)
from Mumbai, a Facebook post with 2.2 million Likes (vividness) by an Indian
lm actor, such as Aamir Khan (culture-boundedness), might have a higher rele-
vance than a similar post by a U.S. lm actor, such as Tom Cruise.
Taken together, for PCs to attract attention and, subsequently, have an inuence
on users’ perceptions, context is necessary. Such context allows plain numbers to
be put into perspective, and it allows users to compare PCs with each other. These
comparisons can be achieved in different ways. First, comparisons can be syn-
chronic or diachronic. While in synchronic situations multiple PCs are available
for direct comparison (e.g., two product reviews presented next to each other),
diachronic comparisons are made when PCs are shown in distinct situations that
occur over time (e.g., when clicking through various products). Hence, synchronic
comparisons are factual comparisons, whereas diachronic comparisons rely on the
users’ memory. Second, comparisons can also be explicit or implicit. That is, while
1 Originally, the concept of salience was only applicable to issues. For the PC context, however,
salience can be ascribed to issues as well as to posts, products, or, broadly speaking, entities.
2 The authors thereby applied the already known psychological concept (e.g., Taylor & Thompson,
1982) to media-effects research.
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
a comparison may rely on the actual PCs of two entities, it can also rely on general
assumptions based on earlier encounters with similar entities. For example, when
being confronted with a New York Times news article and its 723 Facebook Likes,
a user might compare this with either another article containing 512 Likes (ex-
plicit comparison) or with his/her perception of New York Times articles generally
receiving several hundred Likes (implicit comparison).
Currently available research has not adequately dealt with these contextual
characteristics. By subsuming PCs under the umbrella concept of relevance cues,
and by considering the fact that the perception of these relevance cues evolves
over time, we also suggest that users are capable of learning the signicance of
such cues. This assumption is in line with social-learning theory as well as the
literature on perceived self-efcacy (Bandura, 1977). For example, when brows-
ing Facebook every day a user might get a feeling about the value of Likes, and,
thus, be able to differentiate between posts from different originators and their
levels of Likes. Likewise, a tourist that always visits booking.com when planning
a trip eventually ‘knows’ how many ratings provide a reliable forecast for a good
vacation. Thus, variations in both attentional processes and possible effects over
time are expected.
2. Literature review
In light of the theoretical conceptualizations offered above, we now turn to em-
pirical ndings in the domain of PCs by providing a review of academic, peer-re-
viewed studies that examine different general effects of PCs in online media.
2.1 Literature search procedure
All the papers discussed in this literature review were obtained by searching the
following databases: Communication & Mass Media Complete, Web of Science,
Association for Computing Machinery (ACM) Digital Library, and Google Schol-
ar. With one exception, all of the search results were screened; for Google Scholar
results, only ve result pages were taken into account as this database also pre-
sents related hits rather than direct hits, thus increasing the number of results to
an unmanageable amount. The papers had to empirically focus on the effects of
metric user information (e.g., “256 users recommend this book”) to be part of
our sample. Papers that did not focus on metric-related effects of user informa-
tion (e.g., effects of evaluative comments, such as “This book is awesome!”) were
explicitly excluded. To address the problem of conceptual diversity, we dened
two groups of search terms (see Table 1). All reasonable combinations of the
terms within the rst (n=5) and second (n=8) group, such as ‘popularity cues’,
were used to search for potential papers to include in the review. Additionally, the
terms ‘approval ratings’ and ‘rating visualizations’ were included since several of
the papers listed them as additional keywords. During the database search pro-
cess, all the terms were used in quotation marks to enable searching for exact
phrases.
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
Table 1. Search terms used in the literature search procedure
Search Term Group 1 Search Term Group 2
popularity indicators
bandwagon indications
social media bandwagons
user cues
interface information
metrics
ratings
recommendations
Moreover, papers had to have been published between 2005 and 2015. Two rea-
sons justied this chosen time period. First, 2005 was chosen as a starting point
since social-media platforms and so-called Web 2.0 applications started to gain
popularity at this time, thereby also encouraging scientic investigations. These
platforms and applications changed and accelerated the development of recom-
mendation systems, further facilitating the ubiquity of PCs that is prevalent today.
Second, 2015 was chosen as the end-point because we wanted to include research
trends in recent academic discourses on PCs. This was also the reason why we not
only focused on papers in academic journals but also included peer-reviewed
conference manuscripts (full papers only) that tend to be published faster. Pre-
sumably, this enabled us to include papers that reect ongoing research more ap-
propriately.
The initial search yielded 133 unique papers that appeared to be meeting the
access criteria based on the title and abstract. At least to some degree, relevant
papers (peer-reviewed conference manuscripts or journal articles) had to empiri-
cally deal with PCs—dened as metric information about users’ behavior or their
evaluations of entities. Ultimately, 55 articles met our inclusion criteria after
thorough reading (see Table 2). Six articles contained two studies, leading to 61
studies that were quantitatively coded. Due to the large variety of methodological
approaches, and, in some instances, insufcient statistical disclosure, we were un-
able to conduct a statistical meta-analysis of the effect sizes.
Table 2. List of analyzed publications, sorted by name of the author(s)
Author(s) Year of
Publication
Context Method
Ali, Parsons, & Ballantine 2013 E-Commerce and
Marketing
Interview,
experimental
Arora, Arora, & Palvia 2014 E-Commerce and
Marketing
Content analysis,
non-experimental
Bronstein 2013 Online
Communities
Content analysis,
non-experimental
Buder, Schwind, Rudat, & Bodemer 2015 Online News Interview,
experimental
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
Author(s) Year of
Publication
Context Method
Chintagunta, Gopinath, &
Venkataraman
2010 E-Commerce and
Marketing
Content analysis,
non-experimental
Duan, Gu, & Whinston 2008 E-Commerce and
Marketing
Content analysis,
non-experimental
Flanagin, Metzger, Pure, Markov, &
Hartsell
2014 E-Commerce and
Marketing
Interview,
experimental
Fu 2012 Online
Communities
Content analysis,
non-experimental
Go, Jung, & Wu 2014 Online News Interview,
experimental
Ha, White, & Wyer 2012 E-Commerce and
Marketing
Interview,
experimental
Hu & Pu 2014 E-Commerce and
Marketing
Interview,
non-experimental
Jin, Phua, & Lee (2) 2015 Online
Communities
Interview,
experimental
Kelly, Cushing, Dostert, Niu, &
Gyllstrom
2010 Search Engines (Online-)Observa-
tion, experimental
Kim 2014 Online
Communities
Interview,
non-experimental
Kim, Brubaker, & Seo 2015 E-Commerce and
Marketing
Interview,
experimental
Kim & Sundar 2014 Online
Communities
Interview,
experimental
Kim & Sundar 2011a E-Commerce and
Marketing
Interview,
experimental
Kim & Sundar 2011b Online
Communities
Interview,
experimental
Knobloch-Westerwick, Sharma,
Hansen, & Alter
2005 Online News (Online-)Observa-
tion, experimental
Ksiazek, Peer, & Lessard 2014 Online News Content analysis,
non-experimental
Lau, Kwok, & Coiera 2011 Search Engines Interview,
experimental
Lee 2009 E-Commerce and
Marketing
Interview,
experimental
Lee & Jang 2010 Online News Interview,
experimental
Lee & Tan 2013 E-Commerce and
Marketing
Content analysis,
non-experimental
Leino, Räihä, & Finnberg 2011 Online News Interview,
non-experimental
Lim & Steffel 2015 E-Commerce and
Marketing
Interview,
experimental
Luo, Andrews, Song, & Aspara 2014 E-Commerce and
Marketing
(Online-)Observa-
tion, non-experi-
mental
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
Author(s) Year of
Publication
Context Method
Messing & Westwood (2) 2014 Online News Interview,
experimental
Neo 2010 E-Commerce and
Marketing
Interview,
experimental
Nov & Arazy 2015 E-Commerce and
Marketing
Interview,
experimental
Peter, Rossmann, & Keyling 2014 Online
Communities
Interview,
experimental
Porten-Cheé & Eilders 2015 Online News Interview,
experimental
Ringelhan, Wollersheim, & Welpe (2) 2015 Online
Communities
Content analysis,
non-experimental
Rudat & Buder (2) 2015 Online
Communities
Interview,
experimental
Salganik, Dodds, & Watts (2) 2006 Online
Communities
Interview,
experimental
Scott 2014 Online
Communities
Interview,
experimental
Sparling & Sen 2011 E-Commerce and
Marketing
Interview,
experimental
Stavrositu & Kim 2014 Online
Communities
Interview,
experimental
Sugimoto, Thelwall, Larivière, Tsou,
Mongeon, & Macaluso
2013 Online
Communities
Content analysis,
non-experimental
Sundar, Oeldorf-Hirsch, & Xu 2008 E-Commerce and
Marketing
Interview,
experimental
Sundar, Xu, & Oeldorf-Hirsch 2009 E-Commerce and
Marketing
Interview,
experimental
Thuy, Vi, & Linh 2015 E-Commerce and
Marketing
Interview,
experimental
Totti, Costa, Avila, Valle, Meira, &
Almeida
2014 Online
Communities
Content analysis,
non-experimental
Tsay, Dabbish, & Herbsleb 2014 Online
Communities
Content analysis,
non-experimental
Tucker & Zhang 2011 E-Commerce and
Marketing
(Online-)Observa-
tion, experimental
Walther, Liang, Ganster, Wohn, &
Emington
2012 E-Commerce and
Marketing
Interview,
experimental
Weber 2014 Online News Content analysis,
non-experimental
Winter & Krämer (2) 2014 Online News Interview,
experimental
Winter, Krämer, Appel, & Schielke 2011 Blogs Interview,
experimental
Winter, Krämer, Appel, & Schielke 2010 Blogs Interview,
experimental
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
Author(s) Year of
Publication
Context Method
Xenos, Macafee, & Pole 2015 Online
Communities
Content analysis,
non-experimental
Xu 2014 E-Commerce and
Marketing
Interview,
experimental
Xu 2013 Online News Interview,
experimental
Xu, Hao, & Younbo 2015 E-Commerce and
Marketing
Interview,
experimental
Yang 2015 Online News (Online-)Observa-
tion, experimental
Note: Articles containing more than one study are followed by the number of studies in brackets (e.g.,
“(2)” for two studies). The complete listcontaining title and outlet of the publicationcan be
requested from the authors.
2.2 Literature categorization
The categories of the quantitative analysis were derived from literature reviews in
the domain of social media (Kümpel, Karnowski, & Keyling, 2015; Zhang &
Leung, 2015). Due to the given similarity of this research area (online context,
similar ‘key players,’ anonymous yet public communication sphere) the cited re-
views served as a valuable starting point. The derived categories included (a) year
of publication and article type (conference manuscript, journal article), (b) meth-
odological approach (interview, content analysis, observation; each experimental
or non-experimental), and (c) study context (online news, blogs, e-commerce,
search engines, online communities, and marketing). We coded the study context
by assessing the way in which the papers were framed. If a study, such as one by
Messing and Westwood (2014), showed that source cues affect the selection of
online news stories, we coded it within the context of ‘online news.’ Studies, such
as one conducted by Neo (2010), which found that helpfulness ratings had no ef-
fect on purchase intentions, were coded as ‘e-commerce.’
Afterward, we extended the list of categories in order to encompass all aspects
relevant to the investigation of PCs. This extension was based both on the theo-
retical discussion of attentional processes as well as on a qualitative inspection of
the studies, which allowed us to obtain an impression of the research eld. In line
with our focus on media effects, we thus included (d) type of PC (clicks, Likes,
comments, shares, Tweets, favorites, Retweets, rating scales, others), (e) opera-
tionalization of PC extent (e.g., two-digit number for ‘low popularity’), (f)
(in)dependent and moderator variables, and (g) the existence of effects (no effect,
nuanced effect, positive effect).
We categorized PC operationalization by focusing on three dimensions. First,
we distinguished between PCs that are actually deployed (e.g., Facebook Likes,
Amazon stars) and PCs that cannot be found in recent online environments (e.g.,
Facebook dislikes, friendship popularity). Second, we coded the exact types of
PCs (e.g., clicks, Likes, rating scales), allowing for multiple codings if a study fo-
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
cused on more than one type of PC. Third, for all studies that experimentally
varied PCs (n = 11), we also coded the operationalization of the PC extent, indi-
cating what researchers dene as ‘low’ and ‘high’ PC extents. Furthermore, for all
studies that investigated PCs as the independent variable (n = 47), we coded the
reported effects as: (a) none, (b) nuanced, or (c) mostly positive. Moreover, we
coded both the dependent and moderating variables as open-ended variables in a
rst step and re-coded them into categories in a second step. Due to the large
variability and, in some instances, the sheer absence of theoretical grounding, no
category was established to code the studies’ underlying theoretical assumptions.
All studies were read and coded by the authors of this paper in discursive ses-
sions, following three steps. First, all authors of this and Porten-Cheé et al.’s man-
uscript (in this issue) read and coded the studies with a code sheet that included
the categories described above, but, in a rst step, only asked for open-ended
codings. Second, these initial results were discussed and adjusted during a one-
day workshop in early January of 2016. Third, codings were rened and, if pos-
sible, quantied in another round of discursive sessions with all authors of this
manuscript. This procedure called for profound arguments for all codings but
prohibited the calculation of inter-coder reliability.
3. Results
Our sample of articles (N = 55) includes 42 journal articles and 13 conference
manuscripts.3 While only two articles were published prior to 2008, no confer-
ence manuscript from that time met our access criteria (see Table 3). This distri-
bution supports our methodological justication for the chosen starting point of
the investigation. Furthermore, the high number of journal articles published in
2014 (n = 15) and 2015 (n = 10) highlight the current empirical relevance of the
topic in the eld.
Focusing on all studies rather than articles (N = 61), interviews were conduct-
ed in a majority of the studies (n = 42). Most of these interviews incorporated
experimental variations (39). Out of ve (online) observations, four also followed
a post-hoc experimental approach. The remaining 14 studies were content analy-
ses. Overall, PCs were investigated as both dependent and independent variables,
thus allowing for conclusions about the effects derived from PCs and the factors
inuencing PCs. However, the majority of the studies investigated PCs (also) as
the independent variable (n= 52).
3 The two subsamples of journal articles and conference manuscripts do not overlap, except for two
cases: Kim and Sundar (2011) and Winter and Krämer (2014).
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
Table 3. Number of articles investigating PC between 2005 and 2015
Year of Publication
Total2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Article
type
Conference count - - - 1 1 3 3 - 1 2 2 13
Manuscript in % - - - 8 8 23 23 - 8 15 15 100
Journal count 1 1 - 1 1 2 4 3 4 15 10 42
Article in % 2 2 - 2 2 5 10 7 10 36 24 100
Total
count 1 1 - 2 2 5 7 3 5 17 12 55
in % 2 2 - 4 4 9 13 6 9 31 22 102
Note: Discrepancies from 100% in total are due to rounding.
3.1 Contexts in which popularity cues were investigated
The results seem to reect the utilization of PCs within actual online environ-
ments. As such, e-commerce and marketing (n = 23; e.g., shopping websites) and
online communities (n = 20; e.g., social network sites) dominate our sample of
studies (N = 61). Despite this bias, online news (n = 14) clearly prime the rest of
the sample before blogs (2) and search engines (2).
3.2 How popularity cues are operationalized
PCs are operationalized in a wide variety of ways. First, PCs that are actually de-
ployed (rather than PCs that cannot be found in recent online environments) allow
recipients to draw upon the knowledge they gained from prior usage. For example,
if a study is set in a Facebook setting, users of the site are likely to know what Likes
or Shares indicate. While in most studies the second type of PCs is modeled after
actual PCs, users cannot build upon prior knowledge when trying to make sense of
the numbers that are depicted. For example, Hu and Pu (2014) employed both
Likes and Dislikes in an experimental interview where participants were able to
draw upon their prior knowledge about Likes, but they could not build on their
experiences with Dislikes. Out of N = 61 studies, 28 used actual, existing PCs,
whereas 33 used PCs that cannot be found in current online environments. While
studies in the context of online news were equally distributed, the majority of stud-
ies within e-commerce and marketing used non-existent PCs (n=17). The opposite
is true for online communities, where 15 out of 20 studies built on actual, existing
PCs, mostly taken directly from the online community under investigation.
Second, we found a strong tendency toward rating scales within the studies
building on ctitious PCs—out of 33 studies with ctitious PCs, 25 used rating
scales (solely or among other PC types). For example, Lee (2009) investigated the
effects of favorability on a made-up, seven-point rating scale. Real PCs in the
studies mostly incorporated clicks (in 9 studies), Facebook Likes (6), comments
(5), and rating scales which are currently deployed and in use (7), such as ve-star
rating scales (e.g., Knobloch-Westerwick et al., 2005).
Third, for all the studies that experimentally varied the PCs (N = 11), a clear
pattern emerged in which one-digit numbers (n = 7) or two-digit numbers (n = 4)
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
were used as a low PC extent (e.g., 7 Likes, 42 clicks). For the manipulation that
included high PC extents, eight studies used three-digit numbers and three studies
utilized four-digit numbers (e.g., 256 clicks, 1,024 comments). While three studies
reported pre-tests and two studies referred to similar research in order to disclose
how these numbers were derived, six studies experimentally varying PCs did not
justify their operationalization.
3.3 What effects popularity cues imply
The results suggest a Facebook relationship status as a conclusion—“it’s compli-
cated”: While 16 studies found mostly positive effects, 18 reported nuanced ef-
fects, and 13 did not nd any effects. This distribution varies strongly when con-
ference manuscripts and journal articles are examined separately. This indicates a
publication bias toward studies nding (at least nuanced) effects (see Figure1).
Due to the already limited number of studies that consider moderator variables,
and due to the fact that the moderators are usually closely related to the specic
object of investigation, we were unable to deduce quantitative tendencies. Yet, we
offer exemplary insights whenever possible. Overall, our review shows that the
effectiveness of PCs cannot be determined in advance.
Figure 1. Existence of effects by article type
The strongest PC effects were derived from studies within the context of e-com-
merce and marketing (see Table 4). In that context, the affected dependent varia-
bles included further behavior (e.g., intention to purchase), selection (e.g., clicking
on a specic product), and rating (e.g., submitting a star rating after being ex-
posed to PCs). Nuanced effects included PC effects under specic circumstances.
In that regard, the moderator variables include sociodemographic information,
the characteristics of the entity itself (e.g., the weight or look of a product), and
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
the participant’s involvement. For example, in the context of movie selection
decisions, Xu et al. (2015) found that participants who were less familiar with
Hollywood movies relied more on PCs when they made viewing decisions than
participants with higher movie familiarity.
Table 4. Existence of effects by research context
Research Context
Effect(s)
TotalNo effect(s) Nuanced effect(s) Positive effect(s)
Online News 3 3 6 12
Blogs 2 0 0 2
E-Commerce and Marketing 3 4 10 17
Search Engines 2 0 0 2
Online Communities 3 9 2 14
Total 13 16 18 47
Within online news, half of the studies reported positive PC effects. Those studies
included dependent variables, such as selection (e.g., clicking), evaluation (e.g.,
ascribing higher quality), and further behavior (e.g., intention to comment or
share). Nuanced effects within the context of online news are due to the topic and
involvement, need for cognition, and prior knowledge. Such nuanced effects call
for a more differentiated view of PC effects in the context of online news—a rec-
ommendation addressed by Porten-Cheé et al.’s paper in this issue.
In online communities, most of the PC effects are nuanced, including all kinds
of moderator variables, such as a post’s characteristics (e.g., image, headline,
sharing originator), a recipient’s involvement, third-person perception, and need
for cognition. Studying the effects of PCs—referred to as indirect social informa-
tion—Peter and colleagues (2014) investigated the moderating role of the partici-
pants’ perceived importance of PCs. They expected participants that attached
greater importance to PCs to be inuenced more strongly by PCs than partici-
pants that assigned only little importance to them. However, they observed no
such effect.
4. Moving forward in studying popularity cues: Concluding remarks
Scholars from a wide variety of academic disciplines and elds, such as communi-
cation, marketing, social psychology, and economics, have recognized the increas-
ing importance of PCs in online media. PCs provide users with metric informa-
tion about popularity; thus, they help them to make decisions in various
situations, such as when they select or evaluate goods, services, or content. How-
ever, in empirical research, PC researchers have been—and still are—confronted
with strong conceptual and operational ambiguities. By locating PCs under the
umbrella concept of relevance cues, providing a widely applicable denition, and
discussing the attentional processes that lay the foundation for further effects, we
rst tried to decrease this ambiguity and establish a theoretical basis for studying
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
PCs. Second, we analyzed existing empirical research on PCs by conducting a
comprehensive literature review of academic, peer-reviewed studies published
from 2005 to 2015, thus uncovering research patterns and trends in scholarly
activities. Building on this analysis, our literature review suggests that a proto-
typical study on PCs uses experimental surveys to examine the effects of rating
scales on users’ evaluations in an e-commerce setting. It uncovers nuanced effects
prone to moderating inuences, such as a participant’s involvement and an enti-
ty’s characteristics. In the context of our theoretical conceptualization, external
relevance designated by others is apt to affect a user’s evaluation of an entity un-
der certain circumstances. We categorize these moderating circumstances as top-
down vividness and bottom-up salience. For example, PCs may affect users more
when a given product appears to be specic and useful (vividness; e.g., Tucker &
Zhang, 2011). At the same time, PCs have the potential to have a greater effect
on users within e-commerce settings if users are more involved with the purchase
task (salience; e.g., Sundar, Xu, & Oeldorf-Hirsch, 2009).
We acknowledge that a scientic literature review such as ours is naturally
limited by decisions made early in the research process. By choosing to only in-
clude articles that could be found with a predened set of keywords, it is possible
that we omitted research that would also have been relevant for the review. While
we tried to account for publication bias by also including conference manuscripts,
we ignored other sources, such as unpublished papers, dissertations, or research
presented in edited volumes or monographs. Moreover, as this study mostly relied
on a vote-counting approach (Bushman & Wang, 2009), the quality of the stud-
ies, the size of the samples, or the size of the identied effects were not systemati-
cally taken into account. Despite these limitations, we believe that our review
provides useful guidance for researchers. In this concluding section, we seek to
take the results of both the general discussion and the literature review one step
further by providing concluding remarks on current PC research. By doing so, we
offer suggestions on how scholars can move forward in conducting PC research.
Conclusion I: The meaning of PCs has to be learned. The more experience us-
ers have with PCs, the better they are able to use them in their selection and navi-
gation behavior.
Context and implicit or explicit reference points are necessary for PCs to be
effective. This seems especially relevant when conducting experimental research
on the effects of PCs. If researchers do not provide participants with hints on how
given PCs can be interpreted (e.g., by providing explicit points of comparison or
by disclosing which metric values indicate ‘high’ or ‘low’ popularity), participants
are forced to interpret PCs on their own. As most of the experimental studies did
not justify their PC operationalizations, this poses a threat to experimental exter-
nal validity, and it hampers causal inferences. Moreover, it might be interesting to
investigate whether Internet users have already established a sense for interpret-
ing PCs’ extents—regardless of the reference points. It might be helpful for re-
searchers to expose a large number of participants to different types and amounts
of PCs, and ask them about their perceptions. This might uncover the learning
effects induced by repetitive exposure to PCs. Notwithstanding the above, it
seems generally necessary to carefully consider the contextual information that
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
participants use as a basis for their evaluations of PCs—be it in experimental set-
tings or when examining PCs theoretically.
Conclusion II: The effectiveness of PCs depends on external factors, such as
user variables (e.g., informational needs, behavioral intentions, and involvement)
as well as the context variables that determine the vividness and/or salience of
PCs.
While the availability of implicit or explicit reference points is necessary for
PCs to be effective, context alone is not sufcient. Instead, the effectiveness of
PCs depends on the general traits or situational interests and characteristics of the
user. For example, in an online shopping situation, PCs might be irrelevant after
the purchase decision has been made. However, if the user is (virtually) window-
shopping and still in a stage of information seeking, PCs have the potential to in-
uence purchase intentions. Likewise, the users’ level of involvement—reecting
how personally important or interested a user is in buying a product or consum-
ing specic content—moderates PCs’ effectiveness. Decisions that are routine, as
well as decisions that have been made before, are probably far less inuenced by
PCs than decisions a user is unfamiliar or uncertain with.
However, no entity (whether a product or a type of media content) is involving
per se. While buying ofce supplies might be a routine and low-involvement situ-
ation for one individual, another individual—who has had a poor experience
with previous purchases or who is concerned about environmental issues—might
be highly involved in the same situation. Thus, controlling the inuence of in-
volvement seems to be a substantial factor in PC research. Relevant factors that
could inuence an individual’s level of involvement include the availability of al-
ternatives (2 v. 200 available products), the necessity of the decision (mandatory
purchase decision of a new refrigerator v. optional viewing of information, such
as a product description or a news article), and its reversibility.
Our literature review showed that moderating inuences, such as the users’ in-
volvement, their need for cognition, or their prior knowledge, are already included
in some empirical studies on PCs. However, experiments often force decisions on
participants. That is, they make participants choose between alternatives when, in
real-life, a decision would not be mandatory (e.g., “Which of these news articles
would you most likely want to read?”). Thus, a more differentiated view on actual
PC usage situations would strengthen the external validity of PC studies.
Conclusion III: To move forward in PC research, it is necessary to develop a
comprehensive theoretical framework that is open to emerging and evolving on-
line environments.
While we already tried to address the conceptual ambiguity in the eld of PCs
by offering a denition under the umbrella concept of relevance cues, it seems
inevitable to develop a comprehensive theoretical framework. Such a framework
should take attentional processes, (media) effects, and user motivations into ac-
count (e.g., why do people rate movies on IMDB, give Likes for Facebook posts,
or rate products they bought on Amazon?). Put into a scholarly context, a more
recipient-centered approach along the lines of uses-and-gratications research
seems promising and benecial. Moreover, existing models/theories of persuasion
and information processing, such as the Elaboration-Likelihood Model or infor-
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
mational utility approaches, could be used to conceptualize the effects of PCs (see
Porten-Cheé et al.’s paper in this issue).
Ultimately, PC research needs to keep up with recent and rapidly changing de-
velopments in online communication. For example, since Facebook launched its
‘Facebook Reactions’ in February 2016, the variety of PCs—at least in social
media—has increased considerably. Users are still able to ‘Like’ content on Face-
book, but they can also express whether it made them laugh, sad, or angry, and
whether they loved it or were astonished by it. Thus, not only are users able to
easily show reactions that go beyond approval, they can also obtain a better sense
of what others think about a post, a specic type of content, or even a societal
issue. In this regard, PCs might potentially disrupt traditional scholarly perspec-
tives, such as news-value theory. Importantly, PCs might depict relevance toward
recipients as well as communicators (e.g., journalists, e-marketers). Studies have
already shown that, for instance, online newsroom editors “are relying more and
more on digital tracking tools to understand the popularity of news items in or-
der to maximize their presentation of content that audiences will be more likely
to click on” (Lee, 2009, p. 519).
To date, studies have demonstrated the value of PCs as a domain in which to
conduct psychological and social science research—even though much research
still remains to be done. Although we have provided a rst literature review, we
highly encourage researchers to enhance our theoretical and practical understand-
ing of the origins and effects of PCs.
References
Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall.
Bodenhausen, G. V., & Hugenberg, K. (2009). Attention, perception, and social cognition.
In F. Strack & J. Förster (Eds.), Social cognition: The basis of human interaction
(pp.1–22). New York, NY: Psychology Press.
Brosius, H.-B., & Mundorf, N. (1990). Eins und eins ist ungleich zwei: Differentielle
Aufmerksamkeit, Lebhaftigkeit von Information und Medienwirkung [One plus one is
not two: Differential attention, vividness of information, and media effects]. Publizis-
tik, 35(4), 398–407.
Bushman, B. J., & Wang, M. C. (2009). Vote-counting procedures in meta-analysis. In H.
Cooper, L. V. Hedges, & J. C. Valentine (Eds.), The Handbook of Research Synthesis
and Meta-Analysis (pp. 207–220). New York, NY: Russell Sage Foundation.
Duan, W., Gu, B., & Whinston, A. B. (2008). The dynamics of online word-of-mouth and
product sales. An empirical investigation of the movie industry. Journal of Retailing,
84(2), 233–242. https://doi.org/10.1016/j.jretai.2008.04.005
Evatt, D. (1997). The inuence of emotion-evoking content of news on issue salience (Dis-
sertation). University of Texas, Austin, TX.
Galtung, J., & Ruge, M. H. (1965). The structure of foreign news: The presentation of the
Congo, Cuba and Cyprus crises in four Norwegian newspapers. Journal of Peace Re-
search, 2, 64–90. https://doi.org/10.1177/002234336500200104
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb
 Studies in Communication and Media, 7. Jg., 2/2018
Full Paper
Hardmeier, S. (2008). The effects of published polls on citizens. In W. Donsbach & M. W.
Traugott (Eds.), The SAGE handbook of public opinion research (pp. 504–513). Los
Angeles, CA: Sage.
Helmond, A. (2015). The platformization of the web: Making web data platform ready.
Social Media + Society, 1(2). https://doi.org/10.1177/2056305115603080
Hu, R., & Pu, P. (2014). Exploring personality’s effect on users’ rating behavior. In CHI
’14 Extended Abstracts on Human Factors in Computing Systems (pp. 2599–2604).
New York, NY: ACM. https://doi.org/10.1145/2559206.2581317
Kim, H.-S., Brubaker, P., & Seo, K. (2015). Examining psychological effects of source cues
and social plugins on a product review website. Computers in Human Behavior, 49,
74–85. https://doi.org/10.1016/j.chb.2015.02.058
Kim, H.-S., & Sundar, S. S. (2014). Can online buddies and bandwagon cues enhance user
participation in online health communities? Computers in Human Behavior, 37, 319–
333. https://doi.org/10.1016/j.chb.2014.04.039
Kiousis, S. (2004). Explicating media salience: A factor analysis of New York Times issue
coverage during the 2000 U.S. Presidential election. Journal of Communication, 54(1),
71–87. https://doi.org/10.1111/j.1460-2466.2004.tb02614.x
Kisielius, J., & Sternthal, B. (1984). Detecting and explaining vividness effects in attitu-
dinal judgments. Journal of Marketing Research, 21(1), 54–64. https://doi.org/
10.2307/3151792
Kisielius, J., & Sternthal, B. (1986). Examining the vividness controversy: An availability-
valence interpretation. Journal of Consumer Research, 12(4), 418–431.
Knobloch-Westerwick, S., Sharma, N., Hansen, D. L., & Alter, S. (2005). Impact of popular-
ity indications on readers’ selective exposure to online news. Journal of Broadcasting &
Electronic Media, 49(3), 296–313. https://doi.org/10.1207/s15506878jobem4903_3
Ksiazek, T. B., Peer, L., & Lessard, K. (2016). User engagement with online news: Concep-
tualizing interactivity and exploring the relationship between online news videos and
user comments. New Media & Society, 18(3), 502–520. https://doi.org/10.1177/
1461444814545073
Kümpel, A. S., Karnowski, V., & Keyling, T. (2015). News sharing in social media: A re-
view of current research on news sharing users, content, and networks. Social Media +
Society, 1(2). https://doi.org/10.1177/2056305115610141
Lee, Y. (2009). Taking cues from others: The effect of distinct self-views on the persuasive-
ness of extrinsic brand popularity information. Advances in Consumer Research, 36,
981–982.
Messing, S., & Westwood, S. J. (2014). Selective exposure in the age of social media: En-
dorsements trump partisan source afliation when selecting news online. Communica-
tion Research, 41(8), 1042–1063. https://doi.org/10.1177/0093650212466406
Nabi, R. L., & Hendriks, A. (2003). The persuasive effect of host and audience reaction
shots in television talk shows. Journal of Communication, 53(3), 527–543. https://doi.
org/10.1111/j.1460-2466.2003.tb02606.x
Napoli, P. M. (2010). Audience evolution: New technologies and the transformation of
media audiences. New York, NY: Columbia University Press.
Neo, R. L. (2010). Examining the impact of multiple negative online consumer reviews
and review helpfulness ratings on persuasion. Retrieved from http://citation.allacadem-
ic.com/meta/p_mla_apa_research_citation/4/0/3/9/0/p403903_index.html
https://doi.org/10.5771/2192-4007-2018-2-58, am 03.07.2018, 08:05:53
Open Access - - https://www.nomos-elibrary.de/agb

Haim/Kümpel/Brosius | Popularity cues in online media – a review
Nielsen, R. K., & Ganter, S. A. (2017). Dealing with digital intermediaries: A case study of
the relations between publishers and platforms. New Media & Society. https://doi.
org/10.1177/1461444817701318
Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings in social
judgement. Englewood Cliffs, NJ: Prentice Hall.
Noelle-Neumann, E. (1974). The spiral of silence. A theory of public opinion. Journal of
Communication, 24, 43–51. https://doi.org/10.1111/j.1460-2466.1974.tb00367.x
Peter, C., Rossmann, C., & Keyling, T. (2014). Exemplication 2.0: Roles of direct and in-
direct social information in conveying health messages through social network sites.
Journal of Media Psychology: Theories, Methods, and Applications, 26(1), 19–28.
https://doi.org/10.1027/1864-1105/a000103
Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion. Central and pe-
ripheral routes to attitude change. New York, NY: Springer.
Posner, M. I. (1994). Attention: The mechanisms of consciousness. Proceedings of the Na-
tional Academy of Sciences, 91(16), 7398–7403.
Stavrositu, C. D., & Kim, J. (2014). Social media metrics: Third-person perceptions of
health information. Computers in Human Behavior, 35, 61–67. https://doi.org/
10.1016/j.chb.2014.02.025
Taylor, S. E., & Thompson, S. C. (1982). Stalking the elusive “vividness” effect. Psycho-
logical Review, 89(2), 155–181. https://doi.org/10.1037/0033-295X.89.2.155
Walther, J. B., Liang, Y. (Jake), Ganster, T., Wohn, D. Y., & Emington, J. (2012). Online re-
views, helpfulness ratings, and consumer attitudes: An extension of congruity theory to
multiple sources in web 2.0. Journal of Computer-Mediated Communication, 18(1),
97–112. https://doi.org/10.1111/j.1083-6101.2012.01595.x
Weber, P. (2014). Discussions in the comments section: Factors inuencing participation
and interactivity in online newspapers’ reader comments. New Media & Society, 16(6),
941–957. https://doi.org/10.1177/1461444813495165
Xu, X., Hao, X., & Younbo, J. (2015). An information-processing model for audiences’
selections of movies. Journal of Media Psychology, 28(4), 187–199. https://doi.org/
10.1027/1864-1105/a000157
Yang, J. (2016). Effects of popularity-based news recommendations (“most-viewed”) on
users’ exposure to online news. Media Psychology, 19(2), 243–271. https://doi.org/10.
1080/15213269.2015.1006333
Zhang, Y., & Leung, L. (2015). A review of social networking service (SNS) research in
communication journals from 2006 to 2011. New Media & Society, 17(7), 1007–
1024. https://doi.org/10.1177/1461444813520477
... In this regard, Molyneux and Coddington (2020) argue that negative perceptions of clickbait headlines may aggregate over time, whenever a clickbait headline disappoints users, and ultimately, risks losing their trust. Interaction with clickbait may decrease considerably, which is reflected in popularity cues including likes and shares (Haim et al., 2018). Disappointment follows a steep convex negative function (little disappointment is related to strong utility loss), whereas exceeding expected utility is characterized by a flatter concave positive utility function (a lot more utility than expected is required for extra utility gain; Starmer, 2000). ...
... Future research could evaluate if our results hold in countries with more strongly growing online revenues and larger social media audiences, possibly exhibiting greater responsiveness to platforms' regulative intentions. Fourth, likes and shares are certainly of interest to social media editors-although these popularity cues cannot be equated with engagement with journalism (Haim et al., 2018). User interaction in the form of referral traffic might have an even larger impact on media outlets' decisions because this traffic can be directly monetized. ...
Article
Full-text available
Algorithmic curation of social media platforms is considered to create a clickbait media environment. Although clickbait practices can be risky especially for legacy news outlets, clickbait is widely applied. We conceptualize clickbait content supply as a revision game with an unknown threshold. Combining supervised machine learning with time series analysis of Facebook posts and Twitter messages of 37 German legacy news outlets over 54 months, we observe outlets’ behavior following algorithm changes. Results reveal (1) an infrequent use of clickbait with few heavier-using outlets and (2) turning points of clickbait performance as clickbait supply and user interaction form a reversed U-shaped relationship. News outlets (3) collectively adjust toward an industry clickbait standard. While we (4) cannot prove that algorithmic curation increases clickbait, (5) Facebook’s regulative intervention to decrease clickbait disperses heterogeneous tendencies in clickbait supply. We contribute to an understanding of editorial decision-making in competitive environments facing platforms’ regulative intervention.
... Maier 2010; Webster 2014), the concentration of collective attention is accorded increased newsworthiness. Therefore, journalists make use of the growth in accessible usage data and meta-information, which have risen in the course of digitization (Gillespie 2016;Haim, Kümpel, and Brosius 2018;van Dijck and Poell 2013;Webster 2014). This includes publicly visible rankings of most-read or most-shared content, such as trending topics on Twitter (Beckers and Harder 2016;Gillespie 2016;Proferes and Summers 2019;Webster 2014). ...
... Furthermore, metaphors and generalized descriptions are used to represent the participation of a broader public or community (e.g., "storm on Twitter," Beckers and Harder 2016, 913). With the emergence of the Internet and social media, popularity cues and popularity rankings "have become much more common" (Knobloch-Westerwick 2015, 338; see also Haim, Kümpel, and Brosius 2018). Declared trends and the number of shares, likes, and followers can trigger news coverage or can be used to highlight the relevance of a report (Anstead and O'Loughlin 2015;Beckers and Harder 2016;Gillespie 2016;McGregor 2019;Proferes and Summers 2019). ...
Article
Full-text available
News value research assumes that news factors may shift and diversify over time. Despite the technological and economic transformations of the media over the past decades, however, there has been little conceptual work on how journalistic news factors might be changing. This paper addresses how broader developments, such as digitization, datafication, and audience fragmentation, lead to changing news practices. One characteristic of the digital age is that attention is not only scarce and increasingly contested but also continuously measured and compared, leading to a heightened value around everything that seems to draw an exceptional response. Therefore, we introduce the news factor “public response” and argue that journalism increasingly covers those issues and actors that (are said to) have received broad or unexpected public attention. Using numerical, linguistic, and visual means, journalists explicitly tell their audience what many people are paying attention to. We demonstrate the significance of this news factor by pursuing two objectives: First, public response is presented conceptually and distinguished from other news factors. Second, we use the case of the 2016 US election campaign coverage to illustrate public response and derive suggestions for future measurements in qualitative and quantitative textual and visual content analysis.
... To attune to trends in newsrooms, in this study, we consider engagement as a form of interaction, or what Haim et al. (2018) call "popularity cues" as a site where news production and consumption interact. Our goal is to scrutinize audience engagement through aggregated metrics, seeing what newsworkers see, but at a scale that is often out of reach for individual journalists or news organizations. ...
... Our attempts to find "cultures of engagement" that showed explicit patterns of engagement across different national boundaries were based on a broader set of characteristics, such as practices, norms, and values. In that regard, claiming there are cultures of engagement based on three metrics of engagement that could be considered mere popularity cues (Haim et al., 2018) seems like a bet we are not willing to take. However, if we adhere to the reductionist approach that the industry and most often scholarship take toward engagement, and consider likes, shares, and comments as measures of engagement, then yes, we can find distinct patterns of engagement across national boundaries. ...
Article
Full-text available
Information production, dissemination, and consumption are contingent upon cultural and financial dimensions. This study attempts to find cultures of engagement that reflect how audiences engage with news posts made by either commercial or state-owned news outlets on Facebook. To do so, we collected over a million news posts ( n = 1,173,159) produced by 482 news outlets in three Scandinavian countries (Denmark, Norway, and Sweden) and analyzed over 69 million interactions across three metrics of engagement (i.e. comments, likes, and shares). More concretely, we investigate whether the patterns of engagement follow distinct patterns across national boundaries and type of outlet ownership. While we are skeptical of metrics of engagement as markers of specific cultures of engagement, our results show that there are clear differences in how readers engage with news posts depending on the country of origin and whether they are fully state-owned or private-owned outlets.
... Interactive engagement metrics Popularity cues, such as the numbers of likes, comments, and retweets, can serve as an indicator of the success of the media's agenda-setting, framing, and propaganda strategies. These popularity cues have further consequences: they can be used to recommend content on social media platforms and thus impact users' media diets, and they can act as heuristics for people trying to decide what media content is credible, accurate, and important (Haim et al., 2018;Porten-Cheé et al., 2018). Consistent with the idea that the Twitter public sphere is more globallyoriented (especially Western-oriented), independent media posts receive more engagement on Twitter than state-affiliated posts (289.0 vs. 52.7 likes and 69.6 vs. 13.5 retweets, respectively). ...
Preprint
Full-text available
In this report, we describe a new data set called VoynaSlov which contains 21M+ Russian-language social media activities (i.e. tweets, posts, comments) made by Russian media outlets and by the general public during the time of war between Ukraine and Russia. We scraped the data from two major platforms that are widely used in Russia: Twitter and VKontakte (VK), a Russian social media platform based in Saint Petersburg commonly referred to as "Russian Facebook". We provide descriptions of our data collection process and data statistics that compare state-affiliated and independent Russian media, and also the two platforms, VK and Twitter. The main differences that distinguish our data from previously released data related to the ongoing war are its focus on Russian media and consideration of state-affiliation as well as the inclusion of data from VK, which is more suitable than Twitter for understanding Russian public sentiment considering its wide use within Russia. We hope our data set can facilitate future research on information warfare and ultimately enable the reduction and prevention of disinformation and opinion manipulation campaigns. The data set is available at https://github.com/chan0park/VoynaSlov and will be regularly updated as we continuously collect more data.
... For example, many news websites tried to replicate traditional designs, aiming to emulate the newspaper look. Other researchers have looked into these "soft" objects that populate contemporary digital journalism, from the perspective of blogs and hyperlinks (De Maeyer and Le Cam 2015), analytics and metrics (Zamith 2018), popularity cues (Haim, K€ umpel, and Brosius 2018), digital rewards (Ferrer-Conill 2017), or even chatbots (Belair-Gagnon, Lewis, and Agur 2020). ...
Article
Full-text available
This study investigates the visual objects that are used to either disclose or disguise the commercial nature of native advertising as news articles. We adopt a “material object” approach to explore the potential implications for journalism regarding transparency, trust, and credibility. Methodologically, this study used content analysis covering 21 publications in five countries: Germany, Israel, Norway, Spain, and Sweden. We analysed 373 individual native ads. The findings show that news outlets do not follow a consistent way to disclose native ads visually, negotiating the balance between transparency and deception. In this balance, news organizations do not boldly push for transparency and instead remain ambiguous. Our analyses show that both national and organizational characteristics matter when shaping the visual boundaries of journalism.
... In contrast, the majority of the research on interactivity affordances in the political domain has focused on impacts of seeing other users' reactions on oneself (e.g. popularity cues; Dvir-Gvirsman, 2019; Haim, Kümpel, & Brosius, 2018;Walther & Jang, 2012;Yang, 2016) or on the content of conversations between users (Halpern & Gibbs, 2013;Popan, Coursey, Acosta, & Kenworthy, 2019). The current emphasis on SI highlights the potential impact of self-expression on attitude reinforcement. ...
Article
Information is now commonly consumed online, often displayed in conjunction with self-expression affordances (i.e., likes, votes) that create a sense of “self as source.” Sundar et al.’s (2015) theory of interactive media effects (TIME) conceptualizes such affordances as source interactivity (SI). An experiment examined medium effects of SI as well as message effects on attitudes. It tracked selective exposure to attitude-consistent vs. –discrepant political messages, to capture confirmation bias, and manipulated SI presence (affordance to up-vote or downvote articles present or absent) as within-subjects factors. SI use and attitude change were captured. SI reduced selective exposure to attitude-consistent content. However, use of SI affected attitude reinforcement independently as well. Hence, users shaped their own attitudes both by selectively reading articles and expressing their views through SI. Directions for theory development are offered.
... First, the context in which online news is consumed is of great importance as our media system is becoming increasingly hybrid (Chadwick 2013). Previous studies elaborated on the notion of context in many different ways, ranging from different popularity cues (Haim, K€ umpel, and Brosius 2018) to different accompanying user comments (Anderson et al. 2014). In a recent study, Orellana-Rodriguez and Keane (2018) capture the context surrounding the posting of a news-tweet. ...
Article
Full-text available
The complexity and diversity of today’s media landscape provides many challenges for scholars studying online news consumption. Yet it is unclear how news consumers navigate online. Moving forward, we used a custom-built browser plug-in—passively tracking Dutch online news consumers 24/7—to examine how context (website) and content (news topic) features affect patterns of online news consumption. This resulted in a data set containing more than one million Web pages, from 175 websites (news websites, search engines, social media), collected over 8 months in 2017/18. We used automated content analysis to retrieve news topics, and estimated Markov chains to detect consumption patterns. Our findings indicate that news consumers often directly visit their favorite (typically mainstream) news outlet, and continue browsing within that outlet. We also found a strong preference for entertainment news over any other topic. Although social media often offer entertainment news, they are not necessarily the starting point to such news.
... On Facebook, the total number of reactions to content is usually referred to as engagement and includes all types of possible reactions (i.e., likes, shares, and comments). These metrics can be used as an indicator for how popular content was on Facebook and are also subsumed under the term popularity cues (Haim et al., 2018). Although it is contested if popularity cues can serve as an indicator for the approval or positive evaluation of content (and ultimately as a proxy for reputation), they are seen as an indicator for the relevance or salience of content on Facebook. ...
Article
Full-text available
By using social media, corporations can communicate about their corporate social responsibility (CSR) to the public without having to pass through the gatekeeping function of the news media. However, to what extent can corporations influence the public’s evaluation of their CSR activities with social media activities and if the legacy news media still act as the primary agenda setters when it comes to corporate reputation have not yet been thoroughly analyzed in a digitized media environment. This study addressed this research gap by looking at the effect of CSR communication through Facebook and news media coverage of CSR on corporate reputation in Switzerland. The results of this longitudinal study show that the salience and tone of news media coverage of CSR were positively related to corporate reputation, even though the news media coverage about CSR was predominantly negative. Thus, reputation was still strengthened even in the face of negative publicity. No effect of CSR communication through Facebook on corporate reputation was found. The results suggest that legacy news media still were influential in determining how the public evaluates corporations in the digital age.
Article
Full-text available
Zusammenfassung Der Beitrag fragt nach strukturellen Veränderungen der politischen Kommunikation, die sich als Folge der Digitalisierung ergeben. Dabei wird eine regelorientierte und institutionalistische Perspektive eingenommen: Digitale Kommunikationsmedien wie Social-Media-Plattformen weisen eigene institutionelle Logiken auf und beeinflussen so die Regeln, nach denen politische Kommunikation stattfindet. Zur Begründung dieser These wird der Begriff Digitalisierung zunächst in technische Möglichkeit und soziale Realisierung unterschieden. Politische Kommunikation wird als Vermittlungsprozess betrachtet. Neben die Selbstvermittlung durch politische Akteure und die Fremdvermittlung durch journalistisch-redaktionelle Medien tritt mit digitalen Kommunikationsmedien ein neuer Typ in den Vordergrund, die automatisiert algorithmische Vermittlung. Aus diesen Unterscheidungen ergeben sich mehrere Paradoxien, die für eine Betrachtung der institutionellen Folgen relevant sind: Digitalisierung senkt die Kosten der Kommunikation und ermöglicht ein Mehr an publizierten Mitteilungen, erschwert jedoch zugleich die Chance gesellschaftlicher Wahrnehmung und gelingender Kommunikation. Durch die automatisiert algorithmische Vermittlung können Akteure ihre Botschaften in höherer Auflösung an spezifische Zielgruppen richten und sich mit ihnen verbinden, die digitalen Formen der Konnektivität erschweren jedoch die für demokratische Prozesse notwendige Repräsentanz und Zurechenbarkeit von Mitteilungen an politische Akteure. Technisch ermöglichte und sozial eingeforderte Transparenz geht mit der Bemühung von politischen Organisationen einher, das eigene Handeln zu verdecken oder zu verschleiern. Digitalisierung und die automatisiert algorithmische Vermittlung führen damit sowohl zu neuen Sichtbarkeiten als auch zu neuen Unsichtbarkeiten des Politischen.
Article
Full-text available
Social endorsement cues (SEC) offer information about how online users have engaged and evaluated online content. Some view that SEC thus can serve as useful heuristics when users evaluate the credibility of news content on social media. At the same time, SEC can be manipulated by a variety of commercial and political actors on social media. This study examines whether SEC influence individuals' credibility judgments of political news on social media, and how the salience of concerns that SEC can be manipulated by others can undermine the perceived credibility. Using an experiment, we found that SEC had a negative influence on news credibility, regardless of whether or not SEC manipulability concerns were primed. An independent effect of SEC manipulability concerns was also found, such that priming thoughts about the manipulability of SEC led participants to rate the news post as less credible, regardless of whether that post included SEC. These results suggest a spillover effect whereby concerns over the manipulation of SEC can create doubt about the authenticity of other cues from the news (e.g., source and message), and lead to perceptions that news shared on social media can be manipulated more generally.
Article
Full-text available
The rise of digital intermediaries such as search engines and social media is profoundly changing our media environment. Here, we analyze how news media organizations handle their relations to these increasingly important intermediaries. Based on a strategic case study, we argue that relationships between publishers and platforms are characterized by a tension between (1) short-term, operational opportunities and (2) long-term strategic worries about becoming too dependent on intermediaries. We argue that these relationships are shaped by news media’s fear of missing out, the difficulties of evaluating the risk/reward ratios, and a sense of asymmetry. The implication is that news media that developed into an increasingly independent institution in the 20th century—in part enabled by news media organizations’ control over channels of communication—are becoming dependent upon new digital intermediaries that structure the media environment in ways that not only individual citizens but also large, resource-rich, powerful organizations have to adapt to.
Article
Full-text available
As the number of scientific studies continues to grow, it becomes increasingly important to integrate the results from these studies. One simple approach involves counting votes. In the conventional vote-counting procedure, one simply divides studies into three categories: those with significant positive results, those with significant negative results, and those with nonsignificant results. The category containing the most studies is declared the winner. For example, if the majority of studies examining a treatment found significant positive results, then the treatment is considered to have a positive effect. Many authors consider the conventional vote-counting procedure to be crude, flawed, and worthless (see Friedman 2001; Jewell and McCourt 2000; Lee and Bryk 1989; Mann 1994; Rafaeli-Mor and Steinberg 2002; Saroglou 2002; Warner 2001). Take, for example, the title of one article: "Why Vote-Count Reviews Don't Count" (Friedman 2001). We agree that the conventional votecounting procedure can be described in these ways. But all vote-counting procedures are not created equal. The vote-counting procedures described in this chapter are far more sophisticated than the conventional procedure. These more sophisticated procedures can have an important place in the meta-analyst's toolbox. When authors use both vote-counting procedures and effect size procedures with the same data set, they quickly discover that vote-counting procedures are less powerful (see Dochy et al. 2003; Jewell and McCourt 2000; Saroglou 2002). However, vote-counting procedures should never be used as a substitute for effect size procedures. Research synthesists generally have access to four types of information from studies: • a reported effect size, • information that can be used to compute an effect size estimate (for example, raw data, means and standard deviations, test statistic values), • information about whether the hypothesis test found a statistically significant relationship between the independent and dependent variables, and the direction of that relationship (for example, a significant positive mean difference), and • information about only the direction of the relationship between the independent and dependent variables (for example, a positive1 mean difference). These types are rank ordered from most to least in terms of the amount of information they contain (Hedges 1986).2 Effect size procedures should be used for the studies that contain enough information to compute an effect size estimate (see chapters 12 and 13, this volume). Vote-counting procedures should be used for the studies that do not contain enough information to compute an effect size estimate but do contain information about the direction and the statistical significance of results, or that contain just the direction of results. We recommend that vote-counting procedures never be used alone unless none of the studies contain enough information to compute an effect size estimate. Rather, vote-counting procedures should be used in conjunction with effect size procedures. As we describe in section 11.4, effect size estimates and votecount estimates can be combined to obtain a more precise overall estimate of the population effect size.
Article
Full-text available
This article provides a review of scientific, peer-reviewed articles that examine the relationship between news sharing and social media in the period from 2004 to 2014. A total of 461 articles were obtained following a literature search in two databases (Communication & Mass Media Complete [CMMC] and ACM), out of which 109 were deemed relevant based on the study’s inclusion criteria. In order to identify general tendencies and to uncover nuanced findings, news sharing research was analyzed both quantitatively and qualitatively. Three central areas of research—news sharing users, content, and networks—were identified and systematically reviewed. In the central concluding section, the results of the review are used to provide a critical diagnosis of current research and suggestions on how to move forward in news sharing research.
Article
Full-text available
In this article, I inquire into Facebook’s development as a platform by situating it within the transformation of social network sites into social media platforms. I explore this shift with a historical perspective on, what I refer to as, platformization, or the rise of the platform as the dominant infrastructural and economic model of the social web and its consequences. Platformization entails the extension of social media platforms into the rest of the web and their drive to make external web data “platform ready.” The specific technological architecture and ontological distinctiveness of platforms will be examined by taking their programmability into account. I position platformization as a form of platform critique that inquires into the dynamics of the decentralization of platform features and the recentralization of “platform ready” data as a way to examine the consequences of the programmability of social media platforms for the web.
Article
Full-text available
This experimental study (N = 107) aims at investigating how the “most-viewed” recommendation features of online news affect users' news story choices, by employing an unobtrusive measurement for news exposure. The major findings clearly support the thesis that the presence of recommendations influences users' selection of news stories. First, the participants' self-reported assessment of the reasons for their story choices indicates that the incorporation of recommendations heightened their awareness of formal salience features. Second, these recommendations decreased the time for website exploration, and therefore increased the time available for reading news articles. Third, when the recommender system was available, approximately 50% of the participants' total story exposures came through the recommendation features. Fourth, those participants who were in the recommendation condition selected a larger number of the most-viewed stories featured in the recommender system than their counterparts in the no-recommendation condition; and a majority (about 80%) of the former group's access to those most-viewed stories was via the recommender system, in terms of either number or time. Last, the mean exposure time per recommended story did not differ across either recommendation conditions (with vs. without) or access routes (recommender system vs. front/topical sections).
Article
Building on psychocognitive theories of information processing, the purpose of this study was to examine the relative impacts of qualitative bandwagon (e.g., using qualitative information such as comments and reviews) and quantitative bandwagon (e.g., using quantitative information such as view and download counts as bandwagon cues) on content selection decisions by media users. An experiment (N = 77) was conducted to investigate the contributions of the two types of bandwagon behaviors to the selections of Hollywood movies online and to identify factors moderating such bandwagon effects. The results showed that cognitive load was negatively associated with the strength of qualitative bandwagon effects, while positively associated with the strength of quantitative bandwagon effects. Although it was marginally significant, the results also showed that the impact of the quantitative bandwagon effect became stronger when individuals were less familiar with Hollywood movies. Implications with respect to tendency for choice imitation, as well as the conceptual understanding of and methodological approach to the bandwagon behaviors in movie selection, are discussed.
Article
A common belief among marketing practitioners is that increasing the vividness of a message enhances its persuasiveness. This belief has received support in experimental investigations, but vividness also has been found to undermine persuasion or to have no effect. The authors extend a current view of memory operation to predict when and how vividness will affect persuasion. According to this view, the favorableness of available information determines the persuasive effect of vividness. This assertion is tested and supported in a series of experiments. The findings are discussed in terms of strategies for controlling vividness effects.