This article is not the first work to integrate several privacy aspects into a single
structural model. Hui et al. [2006], Li et al. [2010; Li and Santhanam 2008], Xu et al.
[2009, 2011], and Keith et al. [2011] provide similar models which inspired our work.
However, despite the more comprehensive nature of their approach, their work fails
to truly investigate privacy as a decision making process in adequate detail, because
their outcome measure is a more generic form of behavioral intention (i.e., measured
with generic questionnaire items such as “How likely would you provide your personal
information (including your location) to use the M-Coupon service?”). Such intentions
arguably do not directly relate to observable privacy behaviors (cf. Spiekermann et al.
[2001] who show that privacy preferences and actual behavior tend to be weakly related
at best). Our approach, in contrast, considers users’ detailed privacy decisions (a
yes/no decision for multiple disclosures), which is more compatible with existing information
disclosure research (cf. Acquisti et al. [2011]). In effect, our article is arguably
the first to answer the call by Smith et al. [2011] to integrate research on antecedents,
privacy concerns and privacy calculus, with a focus on actual observable outcomes