Chair for Statistics and Data Science in Social Sciences and the Humanities (SODA)
print
Breadcrumb Navigation
Content

Privacy and Synthetic Data

Team-Lead: Anna-Carolina Haensch

Team: Frederic Gerdon, Leah von der Heyde, Marcel Neunhoeffer, Marie-Lou Sohnius

Data Literacy und Evidence Building

Frauke Kreuter and Anna-Carolina Haesch significantly contributed to the design and creation of this interactive training course, a joint venture between NYU Wagner, Accenture, University of Maryland, KYStats, and the Coleridge Initiative. This course offers immersive, hands-on training in data literacy and evidence building.

By utilizing real-world, large scale datasets, participants are given the opportunity to grapple with the application of data in solving pressing societal issues, such as the COVID-19 pandemic. In particular, Anna-Carolina Haesch's synthetic datasets of the KYSTATS datasets is employed as a significant learning tool.

A unique aspect of the course design is its focus on value creation through data. Participants are assembled into teams and tasked with using education and workforce data to construct measures of job quality and education credentials. This collaborative, problem-solving approach not only enhances understanding of data manipulation but also provides participants with a practical, applicable skill set.

https://socialdatascience.umd.edu/data-literacy/

Project team: Frauke Kreuter, Anna-Carolina Haensch

Trust when Sharing Data Online

Decisions about confidentiality protection measures to be applied to data dissemination must be informed by evidence about the utility associated with the quality of the data and the willingness to trade utility against the estimated risk. Doing so requires measurement of data utility, risk, and the willingness of individuals to trade risk for utility. From the theoretical literature on measuring privacy (Nissenbaum 2011) and trust (Bauer and Freitag 2018), perceptions of trust and privacy are context-dependent. There are three dimensions that are particular important: (1) to whom the data is provided, (2) what is done with the data (i.e., whether there are benefits for the one receiving the data vs. benefits for the one providing the data), and (3) what kind of data is shared (i.e., the sensitivity of the data). Some data are inherently sensitive because they touch taboo topics (e.g., information on income, sexual behavior, etc.), other data is only sensitive if it reveals specific information about illegal (e.g., illicit drug use) or counter-normative behaviors and attitudes (Tourangeau and Yan 2007). In this project, we measure utility, risk, and tradeoffs in the context of privacy and data sharing in several cross-sectional surveys. The data landscape has dramatically changed in May of 2018 when GDPR came into effect, and with it the control people have about their data, and the risks companies face when violating GDPR. Thus, we also collect longitudinal data on the awareness about the GDPR regulations in Germany, and in an experimental setting, we measure the influence of GDPR information on trust in various data collecting organizations.

Project team: Frauke Kreuter, Florian Keusch, and Paul C. Bauer

Publications:

The Covid-19 Pandemic and Data Sharing for the Public Good

Does the Covid-19 pandemic lead to a temporary or persistent shift in acceptance of data sharing practices for public benefit purposes? Under which conditions do individuals deem data sharing for the public good as acceptable or even desirable? How can we accommodate data collection and sharing practices and policy to meet the public's preferences? In the project "The Covid-19 Pandemic and Data Sharing for the Public Good: Attitudinal, Ethical, and Legal Approaches to Privacy During the Pandemic and Beyond" we conduct international surveys (in Germany, Spain, and the UK) at multiple time points to measure acceptance of different data sharing scenarios. We investigate how specific situational parameters of the scenarios affect acceptance and how acceptance changes over time, while the purpose of data collection is of particular interest. The project adds to our understanding of the context-dependence of privacy attitudes and, by involving ethical and legal science, informs data collectors and policy-makers on how to design practices and policies such that data for the public good is made possible in an ethical manner. This project is funded by the Volkswagen Foundation.

Project team: Frauke Kreuter, Thomas Fetzer, Helen Nissenbaum, Frederic Gerdon

Publications:

  • Gerdon, F., Nissenbaum, H., Bach, R. L., Kreuter, F., & Zins, S. (2021). Individual Acceptance of Using Health Data for Private and Public Benefit: Changes During the COVID-19 Pandemic. Harvard Data Science Review, (Special Issue 1). https://doi.org/10.1162/99608f92.edf2fc97

The Transatlantic Privacy Perceptions Panel (TAPP)

In the privacy arena, stakeholders from different sectors and with different interests all try to influence privacy debates and resulting policies and practices. The size and diversity of the privacy stakeholder arena present a challenge for systematically synthesizing its members’ conversations and opinions, including reaching, recruiting, and retaining them for a survey.

The TAPP (Transatlantic Privacy Perceptions) Panel aims to follow and analyze developments in stakeholders’ attitudes, expectations, and concerns around current and future topics in digital privacy over time through a cross-national, cross-sectoral survey of influential voices and experts in privacy.

For reaching, recruiting, and retaining this unique population, a novel, tiered and targeted sampling and recruiting approach has been developed, combining systematic research for privacy experts with Twitter-based network and respondent-driven sampling techniques. Informed by mixed-method content analysis as well as qualitative interviews, the TAPP panel will tap panel participants’ thoughts and opinions twice a year in a short survey, including a “privacy index” as a quantitative signal on the state of privacy.

Project team: Frauke Kreuter, Leah von der Heyde, Wiebke Weber

Website: https://privacyperceptions.org/

Pandemic Recovery Survey

The Pandemic Recovery Survey aims to study the impacts of COVID-19 across the world and to assist public health officials in focusing their efforts to recover from the pandemic and allocate resources. It includes questions about people’s access to health information, support, and care, their confidence in vaccines, financial and food (in)security as well as their socio-demographic characteristics. People from 18 countries around the globe are invited to participate via Facebook.

The survey is conducted with the University of Maryland (UMD), the Institute for Health Metrics and Evaluation (IHME) at the University of Washington and Meta Platforms, Inc.

Project team: Anna-Carolina Haensch, Olga Kononykhina, Frauke Kreuter, Wiebke Weber