Designing Repository-Level Evaluation Criteria: A Data Stewardship Framework for Research Data Repositories

Abstract

Purpose: This study aims to design comprehensive repository-level evaluation criteria that assess data stewardship capacity, governance structures, and institutional quality control mechanisms in the context of open science and data-intensive research. Design/methodology/approach: We employed a systematic approach consisting of four stages: formulation of research questions, case study analysis, empirical investigation with expert validation, and framework development through pilot evaluation. We analyzed representative repository-related frameworks and platform-based evaluation guidelines, including the Research Data Alliance (RDA), Confederation of Open Access Repositories (COAR), Science Europe, FAIRsharing, and DSpace, to derive a set of 110 evaluation criteria elements evenly divided between mandatory and optional elements. Findings: The developed framework emphasizes repository-level stewardship, including governance, preservation, provenance tracking, integrity safeguards, and access infrastructure. Rather than directly assessing the scientific validity of individual datasets, the framework evaluates structural conditions that enable trustworthy and sustainable data management at scale. Originality: This research validates and provides comprehensive evaluation criteria for data repositories, filling the gap in previous studies. It creates a balanced framework, distinguishing between essential and additional features.