An ERP Experimental Study About the Effect of Authorization Cue Characteristics on the Privacy Behavior of Recommended Users

Abstract

At the present time, consumers often disclose their privacy when using online platforms to receive personalized recommendation information and services, but they are also highly concerned whether their privacy is being violated. “Privacy paradox” is becoming a hot topic of research. What are the potential impacts of individual cognitive differences and situational cues on privacy decision-making? How to balance the internal causes of the “privacy paradox” so that consumers are more willing to accept personalized recommendation services based on users’ privacy data? Can the transparency of privacy rights ease user conflict perceptions and promote disclosure intentions? These questions are inconclusive. Therefore, the purpose of this our research was to explore consumer privacy paradoxical behaviors from a novel perspective of the characteristics of authorization cues, and to clarify the internal relationship between individual cognitive processing and privacy authorization cues. This study suggests that the big data platform, when collecting or using user information, should try to reduce the behaviors that induce users’ resistance. It also provides a reference for how to better achieve benign interaction in personalized recommendation between Internet companies and users. The event-related potential technique is adopted to explore the matching relationship between individual cognitive processing and privacy authorization cues and to analyze the internal neural mechanism of the personalized recommendation user in the authorization decision. The experiment simulated the privacy authorization situation, and adopted a 2 × 2 × 2 hybrid experimental design: authority sensitivity (high/low) * authorization transparency (with/without permission) * cognitive style (field dependent/field independent). The experimental results show that: (1) Authorization transparency, authority sensitivity and their interactions will affect the user’s privacy authorization behaviors, and the interaction of the two cues has a greater impact on the behavior than the role of a single cue; (2) The cognitive style will affect the individual’s attention resource allocation in the authorization scenario, which, limited by cognitive resources, will result in selective attention to contextual cues: Compared with the field-independent group with self-characterization as a reference, the field-dependent group induced a greater P2 amplitude; (3) When the two-cue valences in the authoritative scenario are inconsistent, the amplitude of the N2 component is greater than that when the valences are consistent, and the amplitude of the N2 induced by the field-dependent group is more affected by the scenario cue valence; (4) Regardless of whether it is a field-dependent group or a field-independent group, there is no salient difference in the amplitude of LPP components induced in each scenario. According to the results of this study, even if privacy authorization involves high risks, individuals tend to selectively seek supportive cues or avoid obtaining information that is inconsistent with their cognition. This research reveals the differences of neural mechanisms in users’ actual decision-making, provides the possibility for further exploration of the black box behind users’ attitudes and behaviors, and opens up new ideas for the study of the “privacy paradox.”