FAIR Social Science Data Enhancement

Research data infrastructures aiming to improve FAIR principles often focus on enhancing metadata. However, most FAIR principles apply to both data and metadata. Interoperability aspects, in particular, can be emphasised, such as standardised knowledge representation, the use of FAIR-compliant vocabularies, and explicit references to other data (Bahim et al. 2020). This presentation discusses areas of data planning and processing that can make data more FAIR. Enhancing FAIRness increases the overall reusability and analytic potential of data, including the possibilities of merging and combining different data sources. It is important to train researchers in established best practices that increase reusability (e.g. http://www.dlib.si/?URN=URN:NBN:SI). For example, comparability of data and conceptual clarity can be improved by using standard social science variables. Standard social science classifications can be derived using shared computer code (see https://www.gesis.org/en/missy/materials/EU-SILC/tools/datahandling). Data infrastructure services are often asked for advice regarding Data Management Planning, which requires maximising FAIR aspects. Historical research projects that use printed archival resources from the past (Mezzoli, 2022, https://doi.org/10.17898/ADP_HDS47_V1) have unified the coding framework from different periods and locations. The variable operationalisations from key social science studies have known reliability and validity characteristics, which the CESSDA European Question Bank, with full survey question text in DDI format, will provide. The use of FAIR-compliant vocabularies, such as the European Language Social Science Thesaurus, to document the conceptual content of data at a granular level can add an additional layer of FAIRness (see https://www.cessda.eu/Tools). In conclusion, a higher return on investment can be expected by publishing FAIR-compliant data (Inau et al. 2023, https://www.jmir.org/2023/1/e45013). Repositories can establish evaluation criteria for data that respect different aspects of FAIR data (e.g. https://pubmet2024.unizd.hr/janez-stebe-abstract/).