Benefits of formalized computational modeling for understanding user behavior in online privacy research

DOIhttps://doi.org/10.1108/JIC-05-2019-0126
Pages431-458
Published date13 March 2020
Date13 March 2020
AuthorTim Schürmann,Nina Gerber,Paul Gerber
Subject MatterInformation & knowledge management,Knowledge management,HR & organizational behaviour,Organizational structure/dynamics,Accounting & Finance,Accounting/accountancy,Behavioural accounting
Benefits of formalized
computational modeling for
understanding user behavior in
online privacy research
Tim Sch
urmann
Work and Engineering Psychology Research Group,
Technical University of Darmstadt, Darmstadt, Germany
Nina Gerber
SECUSO - Security, Usability, Society, Karlsruhe Institute of Technology, Karlsruhe,
Germany, and
Paul Gerber
Work and Engineering Psychology Research Group,
Technical University of Darmstadt, Darmstadt, Germany
Abstract
Purpose Online privacy research has seen a focus on user behavior over the last decade, partly to
understand and explain user decision-making and seeming inconsistenciesregarding usersstated preferences.
This article investigates the level of modeling that contemporary approaches rely on to explain said
inconsistencies and whether drawn conclusions are justified by the applied modeling methodology.
Additionally, it provides resources for researchers interested in using computational modeling.
Design/methodology/approach The article uses data from a pre-existing literature review on the privacy
paradox (N5179 articles) to identify three characteristics of prior research: (1) the frequency of references to
computational-level theories of human decision-making and perception in the literature, (2) the frequency of
interpretations of human decision-making based on computational-level theories, and (3) the frequency of
actual computational-level modeling implementations.
Findings After excluding unrelated articles, 44.1 percent of investigated articles reference at least one theory
that has been traditionally interpreted on a computational level. 33.1 percent of all relevant articles make
statements regarding computational properties of human cognition in online privacy scenarios. Meanwhile, 5.1
percent of all relevant articles apply formalized computational-level modeling to substantiate their claims.
Originality/value The findings highlight the importance of formal, computational-level modeling in online
privacy research, which has so far drawn computational-level conclusions without utilizing appropriate
modeling techniques. Furthermore, this article provides an overview of said modeling techniques and their
benefits to researchers, as well as references for model theories and resources for practical implementation.
Keywords Rationality, Heuristics, Bayesian decision theory, Cognitive modeling, Online privacy, User
behavior
Paper type Research paper
1. Introduction
Academic research investigating user behaviors in online contexts has received substantial
attention in the last decade. To illustrate the rise in research interest, Figure 1 shows the
annual number of publications gathered from Google Scholar concerning the privacy
paradox (Norberg et al., 2007). The phenomenon describes the discrepancy between attitudes
toward ones privacy and actual behavior that would represent these attitudes (Gerber et al.,
2018). The shown research interest led to literature reviews summarizing hundreds of articles
Benefits of
formalized
computational
modeling
431
This work has been co-funded by the DFG as part of project A.2 within the RTG 2050 Privacy and Trust
for Mobile Users.
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/1469-1930.htm
Received 31 May 2019
Revised 5 June 2019
29 September 2019
2 December 2019
Accepted 12 December 2019
Journal of Intellectual Capital
Vol. 21 No. 3, 2020
pp. 431-458
© Emerald Publishing Limited
1469-1930
DOI 10.1108/JIC-05-2019-0126
(Gerber et al., 2018) and detailing a multitude of diverse theoretical explanations (Kokolakis,
2017) for user behavior in online privacy alone.
Online user behavior is not only relevant from an academic perspective. As recent years
have seen several user-data-related scandals for online services, the consequences for public
perception of service providers are currently unclear. However, preliminary research
suggests that negative effects on brand perception are plausible (Hansen et al., 2018). As
customer relations constitute an aspect of an organizations intellectual capital (Luthy, 1998),
avoiding loss of customer satisfaction due to user privacy concerns can be important for
organizations offering online services.
While the term privacy paradoxis nominally tied to online privacy-related user behavior
specifically as opposed to more general online user behavior contexts, its content and scope
appear easily transferrable to the latter. Users of online services tend to express concern
about loss of privacy but still choose to provide online services with private information.
From considering this relationship paradoxical, it follows that humans are generally
expected to act in ways that uphold their preferences. This relationship between preferences
and actions is not specific to online privacy contexts. In fact, the assumption that humans
choose actions that uphold their own preferences in a rational manner is closely tied to
normative decision theory, an interdisciplinary field concerning researchers from philosophy,
economics, psychology, political, and computer sciences, among others (Peterson, 2009).
Thedecision-theoreticalfoundations of researchinto user behaviorare mirrored in several of
the most frequentlyreferenced explanationsof the privacy paradox. A recentliterature review
on research on the privacy paradox (Kokolakis, 2017) establishes, among ot hers, the
predominance of three explanatory approaches: the bounded rationality approach, the
heuristics and biases approach, and the privacy calculus approach. The intuitive notion that
actions should be chosen such that they lead to outcomes corresponding to an individuals
preferences implies the assumption of rationality.The concept of bounded rationality(Simon,
1955),however, describeshow limited cognitiveresources on the side of theindividual may lead
to deviationsfrom this expectation. Applied to privacy-related decision-making, the argument
detailshow limitations in topic knowledgeand computationalcapacities (Kokolakis, 2017)may
limit an individuals potentialfor rational choice. Relatedly,the heuristics and biasesapproach
outlines computational strategies for individuals to adjust to these limitations by using
shortcutsthat routinely delivergood enoughresults,even though they mayresult in behavior
that on occasion does not lead to preferred outcomes. Following this explanatory approach,
neurologically evolved biases in information processing may lead to additional failures to
accuratelyjudge, for example,the risk of privacy loss in anonline interaction (Kokolakis,2017).
0
200
400
600
800
1000
1200
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Publicaons containing "privacy paradox"
Figure 1.
Annual number of
publications
containing the
keyword privacy
paradoxretrieved via
Google Scholar article
search since 2006
JIC
21,3
432

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT