Research funders spend billions on science each year, but they rarely look at how the internal structure of a research group—whether it is hierarchical or flat, large or small, stable or churning—shapes the quality and impact of the work it produces. This matters because most studies of research productivity rely on authorship lists, which conflate who writes a paper with who actually works together day-to-day. The project will combine administrative data from the University of Cambridge and Université de Montréal with bibliometric records and focus groups to map real research group structures—not just publication bylines—and link those structures to output, impact, and novelty. If successful, the findings could help funders and universities design better policies for allocating money and support. For example, a funder might decide to invest in smaller, more collaborative teams in certain disciplines, or to target resources at groups undergoing rapid turnover. The work also aims to develop a method for approximating group structure from publicly available data, so that the approach can be applied widely without requiring access to internal records. This is fundamental social science with a clear practical target: making the research system itself more effective.
View original technical description
Context and topic of research The effective distribution of funds to sustain a productive and healthy research system is a key concern of funders and institutions. To achieve this, governments and research funders commission reviews and set out strategies examining micro- and macro-determinants of research productivity (e.g., CCA 2012, with an updated one in progress; DSIT, 2021; DSIT, 2023). An important, often overlooked factor, in these reviews is the institutional sub-structure and research group structure. For example, the organisational structure of the MRC Laboratory of Molecular Biology (LMB) has been suggested to be a key to producing high-quality scientific output (Nature, 2024). Large-scale quantitative bibliometric studies and small-scale qualitative studies of research collaborations and research group characteristics support the idea that organisational structure can affect the research produced: the relation between team size and citations and productivity (Fox 1991, Lee & Bozeman 2005, Shi et al. 2009); the relation between team size and hierarchy on the disruptiveness of its publications (Wu et al., 2019; Xu et al., 2022); and the relation between gender-diversity and ethnicity-diversity and the novelty and impact of research produced (AlShebli et al., 2018; Yang et al., 2022). Aims and objectives While bibliometric approaches are powerful for exploring collaboration patterns in academia, they conflate authorship lists and research group membership, with the former based on named authors on publications and the latter representing the environment researchers work in. Conversely, this issue is less present in qualitative studies, but they are unable to cover a wide and diverse range of groups. Here, we suggest a unique approach to explore the relationship between organisational structure and research output, by combining administrative, bibliometric, and qualitative data, covering both macro- (across entire institutions) and micro- (individual researchers’ experiences) level information, across two research institutions (the University of Cambridge and Université de Montréal). Unlike previous studies, we will use institutional administrative data to accurately define research groups and their composition and supplement it with bibliometric data to explore how the organisational structure of research groups (devolved/hierarchical, networked/isolated, small/large), structural dynamics (churn and growth rates) and demographics mixtures, relate to the groups’ research outputs, impact, and productivity. Through focus groups with a diverse stratified sample of research groups, we will explore the experiences of researchers in differently structured groups to elucidate how these relate to research production. By comparing internally available administrative data with externally available bibliometric and web-based data, we will develop means to approximate organisational structure using publicly available information allowing us and others to generalise this approach to cases where access to administrative data is not possible. Potential applications and benefits In contrast to previous studies examining authorship lists, this study will reveal the relationship between the structural organisation of research groups and research outputs; and the heterogeneity of these relationships across disciplines. This work will be supplemented by qualitative work to elucidate possible mechanisms underlying these relationships and help to identify key transition points in the structure of research groups which may need particular support. We will convene policy workshops with institutional and national stakeholders to contextualise our findings and develop policy recommendations, which we will publish in a policy brief. Our findings will help policymakers improve the productivity of research by taking into account the effects of research group organisation in different disciplinary contexts and for different objectives.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know