A comparison of different methods of identifying publications related to the United Nations Sustainable Development Goals: Case study of SDG 13—Climate Action

• Published in: Quantitative science studies Vol. 3; no. 4; pp. 976 - 1002
• Main Author: Purnell, Philip J.
• Format: Journal Article
• Language: English
• Published: One Broadway, 12th Floor, Cambridge, Massachusetts 02142, USA MIT Press 20.12.2022; MIT Press Journals, The; The MIT Press
• Subjects: artificial intelligence; bibliometrics; Climate action; Climate change; End users; Identification methods; machine learning; Production methods; Research facilities; Search methods; Strings; Sustainable development; sustainable development goal
• ISSN: 2641-3337; 2641-3337
• DOI: 10.1162/qss_a_00215

As sustainability becomes an increasing priority throughout global society, academic and research institutions are assessed on their contribution to relevant research publications. This study compares four methods of identifying research publications related to United Nations Sustainable Development Goal 13—Climate Action (SDG 13). The four methods (Elsevier, STRINGS, SIRIS, and Dimensions) have each developed search strings with the help of subject matter experts, which are then enhanced through distinct methods to produce a final set of publications. Our analysis showed that the methods produced comparable quantities of publications but with little overlap between them. We visualized some difference in topic focus between the methods and drew links with the search strategies used. Differences between publications retrieved are likely to come from subjective interpretation of the goals, keyword selection, operationalizing search strategies, AI enhancements, and selection of bibliographic database. Each of the elements warrants deeper investigation to understand their role in identifying SDG-related research. Before choosing any method to assess the research contribution to SDGs, end users of SDG data should carefully consider their interpretation of the goal and determine which of the available methods produces the closest data set. Meanwhile, data providers might customize their methods for varying interpretations of the SDGs.