Reducing online identity disclosure using warnings

In an experimental design, we tested whether written warnings can reduce the amount of identity information exposure online. A psychological attack on information privacy that has been shown to be effective in previous research was launched. This attack took advantage of the fact that people respond to certain types of requests in a relatively automatic, or mindless, fashion. The experiment manipulated the word that was used in the alert header: “warning”, “caution”, or “hazard”. All warnings proved to be effective in reducing disclosure, but “hazard” proved to be most effective. Also warnings were more effective in reducing disclosure of driver's license numbers than email addresses. The discussion (a) provides tentative conclusions why these patterns were obtained, (b) suggests how to design warnings in cyber-environments, and (c) addresses future possibilities for research on this topic.