News

Enhancing scientific integrity through reproducibility in Management Science

Publicado em Março 18, 2024

Imagem
Enhancing scientific integrity through reproducibility in Management Science

In a collective effort to enhance the credibility of scientific findings, two recent initiatives have highlighted the significance of reproducibility in the prestigious journal Management Science. Professor Ozkes, associate professor at SKEMA from Digitalization Academy is one of the researchers who contributed to an extensive study on this topic.

​​​​Professor Ali Ozkes​ is an associate professor at SKEMA and Code and Data Associate Editor for Management Science. His research focuses on experimental economics, behavioral game theory, social choice theory, AI ethics and behavior, text-mining, and political economy. Professor Ozkes, alongside his colleagues Miloš Fišar, Be​n Greiner, Christoph Huber, and Elena Katok, embarked on an ambitious endeavor to evaluate the reproducibility of articles published in the journal "Management Science" before and after the implementation of a new Data and Code Disclosure Policy in 2019. 

This project, distinguished by its crowd-science approach, engaged over 700 volunteer researchers from various sub-fields of the journal, dedicating more than 6,500 collective hours to replicate the findings of nearly 500 articles. "Reproducibility in scientific research refers to the ability to obtain consistent results using the same data and methodologies in research articles. This principle is crucial for verifying the reliability of findings and fostering trust within the scientific community and beyond", explains Professor Ozkes.

The acclaimed Management Science Reproducibility Project revealed an impressive 95.3% reproducibility rate for articles compliant with the 2019 Policy when reviewers had access to the necessary data and met the software requirements. However, challenges arose when datasets were proprietary or accessible only through subscriptions, impacting the verifiability of about 29% of the articles. The project underscores the crucial role of data and code availability in ensuring the reliability and integrity of scientific research, extending its implications beyond the confines of "Management Science" to the broader academic community.


Last news