Catarina Pereira
PhD Candidate in Political Science #PoliticalBehavior #ComputationalSocialScience #PoliticalEpistemology
My research examines how discursive, representational, and technological power reshape democracy:
I begin by isolating the democratic consequences of rhetoric, showing how styles like populism and technocracy shape voter attachments beyond policy content. This extends models of public opinion by adding a political system-level rhetorical cue—conceptualized as a property of the party communication system—and linking it to political behavior.
I show that classification-based methods actively produce greater representational bias than distributional approaches, demonstrating that methodological choices themselves shape the political visibility of marginal groups.
Most recently, I explore the impact of AI on political knowledge, showing the epistemic consequences of acquiring political information through conversational interaction.
Across these strands, my work conceptualizes power as a systemic, non-coercive force that structures political behavior, representation, and knowledge: shaping behavior through rhetorical environments, shaping representation through methodological technologies of classification, and shaping knowledge through technological mediation.
My PhD is supported by a Swiss National Foundation project on large scale online deliberation.
You can view my CV updated on May 2025 here
Methods: Machine Learning, Deep Learning, Causal Inference, Instrumental Variables, Text-as-Data
Table of Contents
- Publications
- Grants
- Teaching
- Selected Voluntary Work
- Research Assistant Work
- Education
- Experience
- Data and Resources
Publications
Published Papers
- How to measure political polarization in text-as-data? A scoping review of computational social science approaches - Journal of Information Technology & Politics, 2024. Full View
Working Papers
- Elite Cues and System Legitimacy: How Populism and Technocracy Shape Electoral Competition - C.Pereira, Presented at EPSA and APSA, 2024, 2025.
- How Methods Shape Representation: Classification, Distribution, and the Politics of Measurement - C.Pereira, Presented at CIS Zurichberg colloquium 2025, TADA 2025, Monash-Paris-Warwick-Zurich-CEPR Text-As-Data Workshop 2025.
- Belief Updating through Giving: Conversational AI and Active Epistemic Input - C.Pereira
Grants
- 5 000 CHF: Funding from UZH for a Courses Program in Bocconi University,2025: Advanced Microeconomics and Game Theory courses
- 8 000 CHF: Funding from CIS (UZH + ETH) for Gender Bias Detector, 2024: working paper with Claudia Marangon (ETH)
- 5 000 CHF: Computational Methods Working Group, 2023 CMWG
Teaching
Graduate Level
- Deep Learning For Social Sciences: Teaching Assistant with Prof. Dr. Marco Steenbergen: 2023 2024 [2025]
- Deep Learning For Text and Vision: Teaching Assistant with Prof. Dr. Marco Steenbergen: ESSEX Summer School 2024 [2025]
Ungraduate Level
- Why people don’t vote? Understanding the void: Seminar Syllabus - BA 2024. Access
- Computational Approach to Deliberation: Co-Teaching with Valeria Vuk; Seminar Syllabus - BA 2023. Access
Supervison
- BA Thesis Supervision 2025: Lea Schubarth
Selected Voluntary Work
- European Network Conference 2024: Organizing Committe. Website
- Computational Methods Working Group: Responsible for maintaining the group’s website and organizing workshops CMWG
Research Assistant Work
- Columbia University: Research Assistant for Andreas Wimmer and Prerna Singh (2024-2025)
Education
- PhD in Political Science, University of Zurich, 2022 (Sept) - 2025 (Nov) Swiss National Foundation Project
- Master’s in International Studies, ISCTE - Portugal, 2020-2022
- Master’s in Biomedical Engineering, University Twente - Netherlands, 2014-2016
- Bachelor’s in Biomedical Engineering, IST - Portugal, 2012-2014
Experience
- Data Scientist, Microsoft - EMEA (Spain Based), 2016 - 2020
- Microsoft Research, Hololens Team - Redmond (USA), 2016
- Invited Lecturer of Machine Learning, ISEG - Portugal, 2020-2022