Umeå universitets logga

In the scientific community, authorship is a key mechanism for establishing intellectual priority, securing peer recognition, and assigning responsibility for scientific claims. Over the past two decades, in response to contemporary research practices – characterized by a proliferation of coauthors, increasingly diverse interpretations of authorship roles, and concerns surrounding scientific misconduct – an expanding number of journals have adopted the practice of explicitly disclosing authors’ contributions in published articles.

In this thesis, these author contribution statements serve as the foundation for examining how authorship is attributed, ordered, and valued in the life sciences, a field that has experienced rapid growth in collaborative research. Given the growing reliance on metric-based evaluative mechanisms in academia, the thesis addresses two main objectives: (1) to clarify how authorship is awarded in lab-based life sciences and how authors’ contributions relate to author order, and (2) to validate different models for allocating author credit against the information presented in author contribution statements. This thesis also examines “acknowledgees” – hereafter referred to as subauthors –who are often overlooked despite representing a substantial and sometimes critical segment of the scientific community.

The main results suggest that most authors are core or middle layer contributors, i.e., they perform at least one core layer task or middle layer task. In contrast, most subauthors are classified as outer layer contributors. Authors typically undertake multiple tasks, in contrast to subauthors, and exhibit greater specialization in larger research teams compared to smaller ones. The results also suggest a discrepancy between traditional author guidelines and actual scientific practice regarding author attribution, evident in the significant proportion of outer layer authors.

In addition to these findings, when contributions are ordered by byline position, they exhibit a distinct u-shaped distribution. Among the tested author credit models (u-shaped, arithmetic, fractional, geometric, harmonic), the u-shaped model most accurately reflects observed author contributions, exhibiting the closest alignment with empirical data on core and middle-layer tasks. Furthermore, it also surpasses the other models in predicting which authors are core and which are not.