Intermedial and Multimodal Methods for Researching Public Discourse: How to Combine Qualitative and Computational Approaches

About the seminar:

This week, our guest Dennis Nguyen, Assistant Professor of Digital Literacy and Digital Methods at the Department of Media and Culture Studies at Utrecht University, will present on the topic "Intermedial and Multimodal Methods for Researching Public Discourse: How to Combine Qualitative and Computational Approaches".

Abstract

Computational methods for media data analysis offer novel possibilities for studying the inherently multimodal and intermedial dynamics of mediated communication in contemporary digital culture. However, applying them in valid and reliable ways requires careful calibration of multiple steps, often combining manual and qualitative components within a mixed-methods pipeline. The use of both supervised and unsupervised machine learning methods—for example, in text clustering or image analysis—depends on human input at various stages. The same applies to the deployment of LLMs and generative AI for tasks such as text annotation. Critically, the integration of “human-in-the-loop” approaches in research designs centred on computational methods remains often insufficiently addressed in practice, even though principles of systematicity, transparency, intersubjectivity, and reproducibility have been proposed or can be readily derived from established practices in non-computational methods in the social sciences and humanities. For instance, human validation of automated text clustering and the labelling of clusters can be productively linked to well-established guidelines for content analysis and qualitative coding.

This talk explores the potential of multimodal and intermedial media text analysis, with a focus on key methodological challenges—from data collection and preprocessing to the selection of methods and the reporting of results—illustrated through practical examples centered on public discourse dynamics in media and communication research, including the climate crisis, migration, COVID-19, and rise of AI. It addresses three core questions: (1) what kinds of research questions benefit from a multimodal and intermedial perspective; (2) how to design valid, reliable, and robust methodological frameworks; and (3) what trade-offs, limitations, and advantages computational approaches entail, and why they must be paired with systematic human validation.

Bio

Dennis Nguyen is an Assistant Professor of Digital Literacy and Digital Methods at the Department of Media and Culture Studies at Utrecht University. His research examines public epistemologies of artificial intelligence and datafication, digital cultures and nationalism, crisis discourses, and the role of media in shaping public spheres, as well as the implications of AI for knowledge production. Methodologically, he combines digital and computational approaches with qualitative methods across the qualitative–quantitative spectrum. He also serves as AI Methods Lead at the Inclusive AI Lab, Senior Researcher at Utrecht’s Data School, and affiliate of the Future Literacies Lab. His work has been published in leading international journals, including AI & Society, Big Data & Society, Information, Communication & Society, Journalism Studies, Journalism Practice, International Journal of Communication, and Internet Policy Review. He has presented at major interdisciplinary venues such as IAMCR, ECREA, and ECAI.

How to attend:

To attend the seminar online, please email ims@lnu.se

How to attend:

To attend the seminar online, please email ims@lnu.se