Order Effects in Observations of Stated and Revealed Privacy Preferences

Many Internet services rely on consumers disclosing their personal data. Despite heavy usage and wide acceptance of services like Online Social Networks, doubts about sustainability of trusted relationships remain. Surveying consumers about their preferences reveals severe concerns about the fate of their personal data. In stark contrast to privacy concerns (stated preferences), however, consumers generously disclose personal data in exchange for free Internet services (revealed preferences). It has been argued that individuals experience dissonant states in privacy decision making. The tension between stated and revealed preferences is eliminated with the decision made in order to reduce discomfort. This paper proposes a survey design to determine 1) order effects as indicators for dissonant states in privacy decision making, and 2) the degree of experienced tension between stated and revealed preferences. Observations of data valuation and disclosure behavior are dissonant if they do not commute, i.e. disclosing data prior to valuating privacy does not equal privacy valuation before data disclosure. Determining the degree of dissonance in privacy decision making is expected to inform the design of transparency mechanisms to influence experienced dissonance between stated and revealed privacy preferences.