How do data authors perform in data-intensive research activities? Evidence from author contribution statement in data papers

Abstract

Introduction. Despite the increasing prevalence of data-intensive scientific research, the division of labor in these activities and the performance of data authors remain underexplored. By employing the Contributor Roles Taxonomy (CRediT), this study examines the division of scientific labor in data papers from Data in Brief.

Method and analysis. Utilizing methods of mathematical statistics and data visualization, we analysed the connections between the 14 CRediT roles within data papers. We also explored the relationship between the distribution of labor and the size and discipline of the authorial team, as well as the associations between key authors and their respective CRediT roles.

Results. The results show that 1) data papers rarely make full use of the 14 CRediT roles to describe author contributions. 2) Team size and discipline have a significant impact on the labor division of data-intensive scientific research activities. 3) The need for data collection and analysis is the main reason for the expansion of team size, which is particularly evident in the natural sciences. 4) Corresponding authors and first authors continue to take on core roles. 5) Meanwhile, undertaking data analysis and processing-related tasks, such as ‘Software’, helps authors advance in the author order of data papers.

Conclusions. This study provides insights into the division of labor in data-intensive scientific research and shows that CRediT has limitations in fully capturing the research workflow of data papers. We propose developing a taxonomy specific to data papers, such as DP - CRediT.