Enhancing FAIR compliance: A controlled vocabulary for mapping Social Sciences survey variables

Abstract

The dynamic relationship among survey instruments and study entities like questionnaires, variables, questions, and response formats evolve in Social Sciences surveys. Researchers may need to modify variable attributes such as labels or names, question-wording, or response scales when reusing variables in survey design. Therefore, explaining these relations across different waves and studies is necessary to track how variables relate to each other. Although standards like Data Documentation Initiative – Lifecycle (DDI-LC) and DataCite model these relationships, these frameworks fall short of capturing the complexity of variable relationships. The DDI Alliance Controlled Vocabulary for Commonality Type employs codes—such as 'identical,' 'some,' and 'none'—to outline shifts in entities like variables; however, this approach is insufficient for disambiguating these relationships since they do not differentiate the variable attributes subject to change. We introduce the GESIS Controlled Vocabulary (CV) for Variables in Social Sciences Research Data to bridge this gap. This CV is designed to enhance semantic interoperability across various organizations and systems. Establishing explicit relationships facilitates harmonization across different study waves and enriches data reuse. This enhancement supports advanced search and browse functionalities. The CV, published via the CESSDA vocabulary manager, seeks to forge a semantically rich, interconnected knowledge graph specifically tailored for Social Science Research. This endeavour aligns with the FAIR data principles, aiming to foster a more integrated and accessible research landscape.