Getting the privacy calculus right: Analyzing the relations between privacy concerns, expected benefits, and self-disclosure using response surface analysis

Rational models of privacy self-management such as privacy calculus assume that sharing personal information online can be explained by individuals’ perceptions of risks and benefits. Previous research tested this assumption by conducting conventional multivariate procedures, including path analysis or structural equation modeling. However, these analytical approaches cannot account for the potential conjoint effects of risk and benefit perceptions. In this paper, we use a novel analytical approach called polynomial regressions with response surface analysis (RSA) to investigate potential non-linear and conjoint effects based on three data sets (N1 = 344, N2 = 561, N3 = 1.131). In all three datasets, we find that people self-disclose more when gratifications exceed concerns. In two datasets, we also find that self-disclosure increases when both risk and benefit perceptions are on higher rather than lower levels, suggesting that gratifications play an important role in determining whether and how risk considerations will factor into the decision to disclose information.