If asked, the chances are that most researchers would say that they do not consciously choose to conduct their research with people that are of the same ethnicity as them. They might argue that collaboration choices are driven by considerations including reputation/expertise, previous working relationships, institutional and geographic proximity, funding resources, alignment of research interests or professional network recommendations. However, in a study by Bedoor K. AlShebli, Talal Rahwan, and Wei Lee Woon* the authors delve into the significant role of ethnic diversity in enhancing the impact of scientific research. Through analysis of over 9 million papers and 6 million scientists, the authors explore the relationship between research impact, measured by citation count, and various diversity dimensions, namely ethnicity, discipline, gender, affiliation, and academic age. The findings reveal that ethnic diversity is the most influential factor, showing a strong positive correlation with research impact.
Key Findings
- Homophily in Collaboration: The research identifies a notable tendency for homophily within scientific collaborations, particularly regarding ethnicity, gender, and affiliation. This trend, wherein individuals prefer to collaborate with others similar to themselves, has shown an increase over time, especially in ethnic homophily.
- Ethnic Diversity and Research Impact: Ethnic diversity emerges as the most significant predictor of research impact. Papers authored by ethnically diverse teams receive more citations than those by homogeneous teams. Using coarsened exact matching, the study quantifies that ethnic diversity can enhance a paper’s impact by 10.63% and a scientist’s impact by 47.67%.
- Comparative Analysis of Diversity Dimensions: While ethnic diversity stands out, other forms of diversity such as discipline and affiliation also contribute to research impact but to a lesser extent. This indicates that while interdisciplinary work is beneficial, the diverse perspectives brought by ethnic diversity are paramount.
- Group vs. Individual Diversity: The study differentiates between group diversity (within a paper’s authors) and individual diversity (within a scientist’s collaborators). Both forms positively correlate with research impact, but group diversity has a more pronounced effect.
So, what conclusions can we derive from the study as it relates to EDI – equity, diversity, and inclusivity? The findings underscore the necessity for institutions to foster ethnic diversity within research teams. This can lead to higher-impact publications and a more innovative research environment. Further, the increasing trend of ethnic homophily suggests the need for targeted policies to encourage cross-ethnic collaborations. Finally, incentives for diverse team formations and platforms facilitating such collaborations could be beneficial.
Some potential implications for the IMA include:
- Encouraging Diverse Collaborations: The IMA should promote and support research projects that involve ethnically diverse teams. This could be facilitated through special grants and funding aimed at diverse collaborations.
- Developing Inclusive Policies: Implementing policies to actively counteract homophily, such as mixed-ethnicity mentoring programs or cross-institutional collaborations, will help leverage the benefits of diversity.
- Regular Monitoring and Evaluation: Regular assessments of team diversity and research outcomes will provide valuable insights into the effectiveness of diversity initiatives, helping to refine and improve these policies.
- Educational Programs and Workshops: The IMA can develop educational programs focused on the importance of ethnic diversity in research, aimed at both current members and the broader mathematical community.
By adopting these strategies, the IMA can enhance its commitment to EDI leading to higher-impact research and fostering a more innovative and collaborative scientific community.
What are your thoughts?
* AlShebli, B. K., Rahwan, T., & Woon, W. L. (2018). The preeminence of ethnic diversity in scientific collaboration. Nature communications, 9(1), 5163. https://doi.org/10.1038/s41467-018-07634-8
