interdisciplinarity | Guillermo Romero Moreno

My research interests are in complex systems and wicked problems that require a fundamentally interdisciplinary approach. From the information ecosystems in online social media to the accumulation of long-term health conditions in patients, these problems require a combination of advanced modelling and numerical techniques that are grounded in wide knowledge of the problem domain.

My contribution to the interdisciplinary mix is technical — I employ tools from complex systems, network science, agent-based modelling, and Artificial Intelligence. However, I give considerable care to approaching and understanding the other research fields that are involved. Because models need to be grounded in the phenomena they describe, and only with substantial communication across disciplines this is possible.

Therefore, I see my role as a bridge between communities: understanding, translating, and communicating needs, ideas, and methodologies in both directions. I can bring advanced methodologies to researchers in other fields, and bring their needs and particularities back to inspire new methods.

Common pitfalls in attempting interdisciplinary research

One-sided modelling

Much of the work on opinion dynamics approaches the problem from a physics perspective, often favouring mathematical tractability over realism in how individuals are assumed to behave. Many models live in an “abstract realm”, while the validity of their assumptions and findings depends on context. I believe models should be tied to specific phenomena, situated in time, space, and a social group, and validated either through solid foundations in other disciplines (cognitive science, psychology, social and political sciences) or through empirical data. None of this is possible without working closely with researchers from those fields.

In my collaborations I have repeatedly noticed a gap between how network and statistical methods are understood in theoretical fields and how they are applied in areas such as medicine or political science. Closing that gap requires more than sophisticated methods. It requires translating and communicating them to other communities, making the tools accessible, and listening to what those communities actually need. Something mathematically ‘correct’ and sophisticated may actually not be usable.

Siloed teams

I have also seen the limits of a compartmentalised approach to multidisciplinary projects, where teams from different disciplines work side by side but rarely talk to each other, and meaningful contributions get lost. Interdisciplinarity takes frequent exchanges, proactivity, and genuine effort to read and understand the literature and approaches of other domains.

My interdisciplinarity research experience in practice

Political and social sciences. Together with political scientists, I developed a learning-based method for matching users to parties in voting advice applications. In my PhD, I used agent-based models to study how political party competition shapes consensus and polarisation in the electorate. I have attended the conference European Consortium for Political Research twice, where I have presented my work and engaged with the political science community.
Medicine and healthcare. At the University of Edinburgh, I was part of the AIM-CISC project, a large team of more than twenty researchers from six disciplines (computer science, medicine, sociology, biology, and geography), together with healthcare professionals and members of the public. I worked closely with clinicians to apply network science methods to the understanding the associations between long-term conditions, adapting the methodology to their needs. The work has been published in medical journals and I have presented it to the medical research community.
Cognitive science and psychology. I have worked on opinion dynamics models built on cognitive models of structured beliefs.
Computational Social Sciences. At the AI, Media & Democracy Lab, I work along colleagues from communication sciences, psychology, and law to understand how algorithmic recommendation in social platforms affects disinformation spread and polarising dynamics. We have obtained an interdisciplinarity grant of 10,000€ for our project on how the private message behaviour of young users of TikTok may reduce polarising dynamics.

Building interdisciplinary communities

Interdisciplinary research does not happen on its own. It needs spaces where people from different fields can meet, so I have often taken the initiative to create them:

I organised the AI for Multimorbidity Early-Career Researchers Interdisciplinarity Day, including training on effective interdisciplinary collaboration and peer sharing of good practice, funded by a £10,000 grant from the AIM RSF.
I organised the Symposium on Opinion Shift, Social Influence, and External Manipulation at the University of Edinburgh, with the aim of bringing together researchers who study the topic from different perspective and disciplines. I invited speakers from computer science, psychology, and social sciences, and reached a mixed audience from those disciplines.
I co-chaired the Bridging AI and Psychology (BAIPsy) student workshop, creating a space of exchange for PhD and MSc students working or interested in AI, psychology, and the intersection of the two.
I am currently a IAS Next lead at the Institute for Advanced Study at the University of Amsterdam, where we contribute to maintain a research community for early-career researchers working on interdisciplinary topics.
I co-founded and co-organised a cross-departmental Networks Reading Group at the University of Edinburgh.

Looking ahead

I advocate for a more interdisciplinary scientific approach, and I recognise that the institutional context is essential for it to succeed: interdisciplinary work needs time, funding, and recognition that traditional academic incentives do not always provide. My aim is to keep integrating my methodological expertise with the domain knowledge of researchers from other disciplines, to validate models against real phenomena, and to increase the significance and impact of research beyond the boundaries of any single field.