

45
Saturday, November 12
1 6 : 0 0 – 1 7 : 3 0
ARS17
Audience Research in the Age of Networks
PP 675
The Filter Bubble and Its Consequences: A Longitudinal Field Experiment
C. Courtois
1
, C. Lennert
1
1
KU Leuven, School for Mass Communication Research, Leuven, Belgium
Web search has grown to be a fundamental media practice in handling the abundance of online information (Couldry, 2012). The affording algorithms have
become increasingly powerful, partly due to their high degree of personalization. That is, search results are based upon prior individual online behaviour.
Pariser (2011) argues that current personalization algorithms lead to highly selective exposure to information that aligns with users initial viewpoints.
Hence, based on a user’s online behaviour, counter-arguing information is filtered out, leaving beliefs unchallenged while browsing in a self-created in‑
formation bubble. On a societal level, this phenomenon potentially impedes public debate and the conflict of discourses that underlies it. Still, it is unclear
whether such bubbles have an impact. Following Hjarvard’s view on Mediatization (2013), we conceptualize this phenomenon through the lens of Giddens’
Structuration Theory (1984). More specifically, we consider online search as intersecting practices tied to search engines as social structures, which are pro‑
duced and reproduced by the interplay of technology companies and internet users. Our study focuses on Google Search, hypothesizing that varying degrees
of personalization – either self-induced or imposed by the search platform itself – affect beliefs, attitudes, and behavioural intention in the field of health,
politics, economics and ecology. To test this central hypothesis, a double blind longitudinal field experiment was set up. At the beginning of the study,
a panel of 400 social sciences students registered for a new Google account and filled out a questionnaire on media and search behaviour and ecological,
political, economic and health attitudes and behaviour. During the experiment, which lasts for two months, students were asked to search while logged in
to their accounts in exchange for a course credit. The panel was randomly divided in three conditions: (a) control group personalizing at its own pace, (b)
an experimental group A whose accounts are trained automatically to search for information supporting a set of predefined beliefs in the realm of politics,
ecology, economic and health, and (c) an experimental group B whose accounts are trained to search disproving information for the same beliefs. During
the study, all participants’ degree of personalization is measured automatically by means of a script that uses the participations access data to emulate
Google Searches directly through its API. After one or two months, the participants are invited once for a lab experiment. During this experiment, beliefs in
the fields of health, politics, economics and ecology are presented, asking the participants whether or not they think these are true, and how confident they
are of their answers. Next, the participants are invited to freely search Google, either logged into their account (i.e. the control group or experimental groups
A or B) or not logged in at all. During this free search, participants can revise their initial answers and update the confidence therein. Moreover attitudes and
behavioural intentions are re-assessed. During the analysis, these measures are contrasted with prior answers, considering their type of personalization.
The study is ongoing, expecting the final results to be analysed by June 2016.
PP 676
Growing Trust in Facebook on the Smartphone (in Norway in 2012)
L. Nyre
1
, B. Tessem
1
1
University of Bergen- Norway, Department of Information Science and Media Studies, Bergen, Norway
In Norway in 2012 mobile apps for social media were being diffused. The early majority of people in this Scandinavian country had started using Facebook
on the smartphone, but a large proportion of the population had still not adopted the new practice. This study addresses the transformation that happens
when a person adopts a new media practice like using Facebook on the mobile. Our argument is based on a research project (N=99) in Sogndal, Western
Norway where we found that the more informants used social media on the mobile; the more they trusted social media with their personal information.
People shared personal information despite risks of unwanted public exposure, privacy breaches due to tracking, surveillance, etc. The readiness to trust
Facebook on the smartphone in Sogndal in 2012 is discussed as a historically contingent attitude, and it is analysed in detail distinguishing between in‑
formants' relations to other persons, societal institutions and social media. The paper ends with a critical discussion of the informants' propensity to trust
Facebook on the smartphone with personal information. Is Facebook in this case functioning as a (parasitic) placeholder for trust that would have been
better directed elsewhere? Keywords: smartphone, mobile social media, Facebook, trust, confidence, Norway
PP 677
“It Could Happen That...” An Analysis of Perceived Risks of Smartphone Applications
J. Wimmer
1
, M. Neunkirchen
1
1
TU Ilmenau, Institute for Media and Communication Science, Ilmenau, Germany
Over the years, our everyday lives have become intertwined with digital media. Empirical studies document an increase in usage duration and intensity,
as well as a high level of routine usage for all existing mobile technology (from smartphones and tablet computers to hand-held consoles), which is not
always reflected by the users. The attractiveness of mobile media is tightly linked with the amount of trust a consumer places in it. Trust therefore forms
a key component from the audience perspective in the postmodern era, and on the other hand is fundamentally linked to social and media change. While
research into the risks of the digital media world is concentrated on specific applications, such as Facebook, investigation into the broader usage and
repertoire of apps and their growing complexities is still lacking. The proposed analysis therefore explores which risks app users are conscious of and which
factors influence this risk awareness. The first phase of the study consisted of a quantitative online survey based on theoretical considerations, which was
posted on all important German technology related forums, as well as various email-distribution lists, in order to reach a wide range of app users. Focusing
on the highest possible heterogeneity from this pool, thirteen people, primarily mobile-operating systems users, were invited to an interview for the second
step. The guided interviews solidified a total of three dimensions of perceived risks similar to the state-of-research on consumer behaviour in E-Commerce:
Financial, technological and psychological risks. The changed contexts of mobile media ecology also lead to another perception of risks. Interestingly,
the interviewees did not mention social risks, such as social disadvantage or exclusion from ones peer group as a result of not using certain apps. Although