Bits and Pieces: Persistent identifiers (PIDs) for dataset elements.

We present the critical role of PIDs in enhancing data citation and reusing research data at granular levels. It highlights the efforts of the KonsortSWD consortium, part of Germany's National Research Data Infrastructure (NFDI), in implementing the PID Registration Service. The discussion focuses on PIDs as stable, unique references that address challenges in data citation, such as dataset versioning and accessing specific elements like variables, questions, and excerpts. The presentation outlines the development of the PID registration service, designed to assign PIDs below the study and dataset level, emphasizing its application in diverse use cases within social sciences and economics. Additionally, it introduces knowledge graphs and controlled vocabularies to map relations between variables across studies and waves, fostering semantic richness and interoperability. The talk concludes with practical benefits for researchers and data providers, along with future directions to extend PID services to other data types, such as qualitative data.