| Sumario: | Since the advent of social network sites, researchers have investigated how and why users share personal information online. Yet, the replicability of individual findings remains unclear. We addressed this gap by closely replicating three seminal studies: Krasnova et al.'s (2010) study on the privacy calculus, Vitak's (2012) analysis of the impact of context collapse, and Dienlin and Trepte's (2015) investigation of the privacy paradox. While only 32.5% of the original effects replicated exactly across the three studies, 67.5% were significant and in line with the original direction. Despite this overall replication success, the prominent negative privacy concerns and self-disclosure link did not replicate and became positive instead. Additional specification curve analyses revealed that the strength of this relationship is contingent on a variety of analytical decisions. The findings offer insights into the replicability of survey-based privacy research, highlighting the role of replication in a dynamic research landscape.
|
|---|