Synthetic datasets for replication and teaching purposes

Abstract

Many research applications in the social sciences require a high degree of confidentiality to protect research subjects. This hinders the need for data access to other scientists and open access standards and workflows. This is particularly true for qualitative data sources as part of Mixed Methods research designs. Synthetic datasets feature the same statistical characteristics of original datasets, while making a traceback to research subject identities impossible. This presentation briefly presents the methods involved in creating synthetic datasets and some suggestions, how such synthetic datasets could be used to allow for Open Access and transparent data access while also maintaining the required high ethical and legal data protection standards required for sensitive data collections. It also provides some ideas about using synthetic datasets as part of methods teaching.

Joachim K. Rennstich

Professor International Social Work and Research Methods

My research interests include the long-term development of digital capitalism, digital literacy and innovative teaching-methodologies.