Figure 1.1. Interactive Forces in Tension. Source: Modified from Neuman 1991.

Figure 2.1. Digital Transformation of Business Model

Figure 2.2. Digital Transformation of Audience Measurement. Note: Arrow denotes confidence range of audience-user measurement.

Figure 2.3. Efficiency Gap in Marketplace Performance. Note: Denotes the degree of interaction between privacy and surveillance denotes the degree of personalization. Here a cutoff point for low and high is only conceptual to illustrate the interaction.

Figure 2.4. Marketplace Performance of Privacy Protection: Notice and Choice. Note: Percentage of the sampled websites shown.

Figure 2.5. Marketplace Performance of Privacy Protection Over Time. Note: Percentage of the sampled websites in each study shown.

Figure 2.6. Marketplace Contrast between Privacy Consumption and Production. Note: For privacy protection, percentage of the sampled websites (Park 2008 and LaRose 2002) shown. See Appendix A for various sources for privacy concern.

Figure 3.1. Distribution of Privacy Knowledge Items

Figure 3.2. Moderating Effects of Privacy Knowledge. Note: Privacy concern is informational, with the behavior indicating the level of privacy control in technical dimensions (Appendix B). *p < .05. Source: Knowledge Network data.

Figure 3.3. Direct Effect of Willingness for Privacy Trade-Off. Note: Willingness is the measure of the extent to which an individual is willing to trade personal data for rewards, with the behavior denoting privacy control in the technical dimension (Appendix B).

Figure 3.4. Acquisition of Privacy Knowledge according to Sociodemographic variables. Note: For KN sample, entries are odd ratios. The odds larger than 1 indicate the likelihood of the correct responses. Covariates (yearly experience; number of online accesses; daily use) are not shown in logistic regression. Policy understanding is an item asking about appropriation. Surveillance awareness is tone asking about transfer. Solid lines are for surveillance awareness; dotted ones are for policy understandings. For mobile sample, covariates (mobile familiarity; mobile access) are not shown in multivariate regression. Entries are standard coefficients. * p < .05; ** p < .01; ns = nonsignificant.

Figure 4.1. Privacy Policy Orientation. Note: X = Emphasis on well-being, as opposed to business interest.welling between business and citizens. Y = Policy preference between market-driven and privacy-driven. Source: Modified from Dutton and Peltu 1996.

Figure 5.1. Personal Data Ecosystem in Vertical and Horizontal Concentration

Figure 5.2. Data Submission (Person) and Use (AI) as a Prisoner’s Dilemma. Note: A = value for digital platform, such as advertising, target ad, microcustomization, and marketing. B = value for a person, such as access and customized and automatic suggestions. CA = cost for AI, the opportunity to microtarget a person. CB = cost for person, AI-generated automation.

Figure 5.3. Emotional Microtargeting in Automated Shopping

Figure 5.4. Political Microtargeting in Facebook–Cambridge Analytica

Figure 5.5. AI-Based Facebook Effect Metrics versus Mass-Media Effect Model. Note: A dotted line indicates potential connection. For instance, media buyers often engage in microtargeting as well as a mass-media campaign.

Figure 5.6. Sequence of AI Decision Error: Types I and II

Figure 5.7. Logistic Growth Pattern in AI Return of Data

Figure 6.1. Regulatory Codes for Privacy over Surveillance. Source: modified from Lessig 2009.