Beyond Monkey Jobs. Leveraging a 'Data as Code' Approach for Efficient and FAIR Research Data Management
Scisci, Domingo; De Santis, Giovanna
- Publisher
- Zenodo
- Language
- en
Abstract
The growing complexity of research data infrastructures presents both opportunities and challenges for social science research. A persistent hurdle is the burden of repetitive, error-prone data management tasks that consume valuable time and resources. The ‘Data as Code’ (DaC) paradigm offers a transformative solution by applying software development methodologies to research data management. This approach enhances reproducibility, efficiency, and transparency, aligning with the Open Science and the FAIR (Findable, Accessible, Interoperable, and Reusable) principles for digital research infrastructures, thereby fostering methodological rigor and innovation. This presentation explores the integration of programmatic tools, artificial intelligence, and the DaC approach within social science data infrastructures, contributing to ongoing discussions on the future of research data management while providing practical insights for researchers and data archivists to improve data usability and long-term preservation. Through case studies from the Data Archive for Social Sciences in Italy (DASSI)—the Italian Service Provider of CESSDA ERIC—we demonstrate how these methodologies improve data quality, reproducibility, and long-term preservation. By reducing manual interventions and enhancing data curation processes, this approach minimizes human error and bolsters data archiving, accessibility, and interoperability, thereby allowing professionals to dedicate more time to critical, high-value tasks. By embracing the Data as Code paradigm, research infrastructures can move beyond traditional matrix-based data structures, enabling seamless integration and management of diverse data formats.This flexibility ensures that heterogeneous datasets—ranging from hierarchical and nested data to unstructured sources—can be seamlessly integrated, curated, and disseminated, expanding the scope and potential of social science research.
Full text
Beyond Monkey Jobs Leveraging a ‘Data as Code’ Approach for Efficient and FAIR Research Data Management Domingo Scisci, Giovanna De Santis ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology NAPLES - 09 | 23 | 2025
ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology - Naples - 23 | 09 | 2025 The Challenge Increasingly large volumes of data more complex management Reproducibility and transparency as key values How to ensure them? Need for more and more structured (meta)data F & I of FAIR principles Manual, repetitive tasks Time-consuming and error-prone
ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology - Naples - 23 | 09 | 2025 What is Data as Code? TRADITIONAL DATA AS CODE data_ok.dta script.do script1.do data_def.dta data_ok_ok.dta data_def_latest.dta script2.do script2.do import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() main.py import/ analysis/ CODEBASE import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() import pandas as pd def echo_world() { print("hello_world") } if __name__ == "main": echo_world() Focusing attention (and work) on the code that generates the data as opposed to the data itself, facilitating activities to document the operations performed, in a transparent and collaborative manner. More datasets around: which one is updated? Self-contained scripts: how to deal with conflicts? Single Source of Truth Focus on code Always updated for all
ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology - Naples - 23 | 09 | 2025 Key Features of Data as Code VERSION CONTROL AUTOMATION REPRODUCIBILITY quarto
ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology - Naples - 23 | 09 | 2025 Case Study: DASSI Data Archive for Social Sciences in Italy Led by the University of Milano-Bicocca and CNR National Service Provider of CESSDA ERIC Main services: Data curation and archiving for long-term preservation Data discovery and access Support for researchers in data management Many activities require repetitive tasks and careful documentation, which are inherently prone to error and thus demand rigorous quality control.
ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology - Naples - 23 | 09 | 2025 Example 1: Data Management - ITA.LI
ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology - Naples - 23 | 09 | 2025 Example 1: Data Management - ITA.LI
ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology - Naples - 23 | 09 | 2025 Example 1: Data Management - ITA.LI
ISA - RC33 Logic and Methodology - 11th International Conference on Social Science Methodology - Naples - 23 | 09 | 2025 Example 1: Data Management - ITA.LI
THANK YOU! Questions? Domingo Scisci [email protected] domingoscisci Giovanna De Santis [email protected] © 2025 Domingo Scisci, Giovanna De Santis. This work is licensed under a CC BY 4.0 International License
