A multi-strategic approach to locating institutional data deposits

Abstract

Fragmentation in the landscape of data sharing poses a challenge to institutional attempts to assess research output and compliance with grant requirements. Variation in disciplinary norms, individual practices, and metadata standards make it difficult to determine where, or whether, affiliated researchers deposit their datasets at project’s end. While commercial providers do offer paid services purporting to solve these issues, many institutions prefer to keep this work in-house for reasons of cost, accuracy, and fit with local objectives. Practical examples of dataset discovery projects, their successes and failures, are useful to decision-makers within research institutions weighing their options in this area. We report the results of a search for publicly accessible datasets produced by Auburn University researchers within the past ten years. This effort, a collaboration between two librarians and an undergraduate student, involved multiple strategies for finding data. Examples of such included searching databases for publications with associated data, searching generalist and specialist repositories directly, and utilizing the scholarly profiles (e.g. ORCID records) of known researchers of interest. Lists of federal grants received by Auburn-affiliated researchers were used to help identify and prioritize potential data depositors. This poster explores the rationale behind the various strategies along with their relative success in discovering data. As the purpose of the project was to iteratively develop a methodology, the roadblocks we encountered were as instructive as the successes of each strategy. We conclude with a discussion of planned next steps, such as discussions with institutional stakeholders, and a broader reflection on the challenges posed by inconsistent or absent affiliation metadata in the research reporting infrastructure.