From Antagonistic Talk to Cooperative Acts: The Effect of Polarization Cues on Newspaper Audiences
Two laboratory experiments test whether news about a polarized society can itself change polarization-related behavior.
Article ↗Computational social science · public communication · behavior
Mike Farjam
Integrating computational evidence across levels of society
I study how public communication makes societal conflict visible, how people perceive that conflict, and when those perceptions shape behavior. My work combines computational text analysis and AI with digital traces, surveys, and experiments.
Add assets/photo.png
Profile
Different methods. One evidential architecture.
I am a Senior Research Associate in communication studies at the University of Hamburg. My academic background spans artificial intelligence, psychology, economics, and communication science. That breadth shapes how I approach computational social science: not as a single toolbox, but as the task of connecting different forms of evidence to the same social process.
My current substantive focus is public communication under conditions of societal conflict. I am particularly interested in polarization, climate and protest communication, digital media, and the ways journalistic and political representations influence perception and behavior.
Selected work
Recent publications
Recent work spans media effects, computational communication research, LLM-assisted content analysis, multilingual digital discourse, and the public representation of polarization.
The polarized talk least about polarization
A comparative computational analysis of polarization discourse in British, German, and Swedish newspapers.
Article ↗A Practical Guide and Case Study on How to Instruct LLMs for Automated Coding During Content Analysis
A validation-focused workflow for using locally run large language models in social-scientific content analysis.
Article ↗The re-mediation of legacy and new media on Twitter
A six-language study of the European migration discourse, connecting media-source patterns, topics, and toxicity on Twitter.
Article ↗Assessing Threat and Efficacy
A comparative study of ideological differences in climate-change reporting in US and German newspapers from 2006 to 2023.
Article ↗Do you really believe that?
An incentivized experiment on the acceptance of real-world evidence in politically polarized domains.
Article ↗Projects & funding
Research in collaboration
My current work combines computational research with experiments and direct collaboration with journalists. I have experience coordinating international and transdisciplinary projects and linking methodological work across institutions.
2025–2028
Media and Polarization in Europe: How Journalism can Support Democratic Debates
Volkswagen Foundation · Co-applicant and project coordinator
International comparative project across Germany, Sweden, Spain, and Poland. We study how journalistic practices relate to public perceptions of polarization and work with journalists to develop evidence-based recommendations for democratic debate.
Podcast · In German
Medien zwischen Polarisierung und Meinungsbildung
A conversation with me about our current research on media and polarization.
Listen on Spotify (opens in a new tab)
2024–2025
Political Polarization and Journalistic Practices: Adding Fuel to the Fire?
Volkswagen Foundation · Co-applicant and project coordinator
The pilot project that established the Fuel2Fire research program, combining media-content research, experiments, and exchange with journalistic practice.
2021–2023
PROTECT – The Right to International Protection
European Union Horizon 2020 · Senior Researcher, Lund University
Research on migration discourse, including large-scale multilingual analysis of social-media communication and the interaction between legacy and new media.
2018–2020
Modeling institutional dynamics in commons
Riksbankens Jubileumsfond · Co-applicant
Interdisciplinary work combining historical data, institutional analysis, computational modeling, and agent-based simulation.
Teaching
From questions to evidence
My teaching follows the sequence question → evidence → data → method → inference. I emphasize quantitative foundations, transparent workflows, and the ability to judge what a method can and cannot establish.
2,300+
contact hours of university teaching at Bachelor's, Master's, and PhD levels
I hold 15 ECTS in Higher Education Teacher Training and 15 ECTS in formal postgraduate-supervision training. I have supervised 12 Master's theses and currently co-supervise a PhD student.
Experimental Political ScienceUniversity of Hamburg
Bayesian StatisticsUniversity of Milan
Quantitative Data Analysis in RLund University
Quantitative Methods I & IILinnaeus University
Experimental MethodsLinnaeus University
Programming in NetLogoESSA Summer School
Academic trajectory
AI, psychology, economics, communication
My interdisciplinary trajectory is the basis for my current research identity: computational methods matter most when they are connected to theory, measurement, and causal evidence about social processes.
2024–present
University of Hamburg
Senior Research Associate · Communication Studies, Climate and Science Communication
2021–2023
Lund University
Senior Researcher · Horizon 2020 PROTECT
2016–2020
Linnaeus University
Postdoctoral Researcher → Senior Lecturer
2013–2016
Friedrich Schiller University Jena / IMPRS
PhD in Economics, summa cum laude
Earlier
Radboud University Nijmegen
M.Sc. & B.Sc. Artificial Intelligence · B.Sc. Psychology
Contact
University of Hamburg
Faculty of Business, Economics and Social Sciences · Social Sciences · Journalism and Mass Communication
Von-Melle-Park 5 · Room 210320146 Hamburg · Germany