From Posts to Insights: Navigating Social Media Data Challenges in Research

10:30 - 11:00     Walk-in Coffee, Tea & Registration

11:00 - 11:05     Welcome and opening

11:00 - 11:20     Opening remarks by Jeroen Sondervan (Open
Science NL)

11:20 - 11.30     Introduction: Social Media Data Networking Event

11:30 - 12:00     Community engagement session: Mapping roles and challenges

12:00 - 12:30     Keynote speaker: Tim de Winkel (University of Applied Sciences Utrecht)

12:30 - 13:30     Lunch Break

13:30 - 14:30     Panel discussion: Legal & Ethical Foundations: 
                                  Carola Veenstra - Jager (University of Groningen),
                                  Paddy Leerssen (University of Amsterdam)
                                  Roel Lutkenhaus (University of Twente)
                                  Giovanni Spitale (Universität Zürich)
                                  Moderator: Efe Sözeri (University of Twente)

14:30 - 15:00     Coffee break

15:00 - 16:00     Parallel Session
                                  Session 1: Giovanni Spitale (Universität Zürich)
                                  Session 2: Karel Kroeze (University of Twente)
                                  Session 3: Danielle McCool (University Utrecht)

16:00 - 16:30     Keynote speaker: Jessica Taylor Piotrowski (University of Amsterdam)

16:30                 Closure + Networking Drinks

12:00 – 12:30 Keynote speaker

Tim de Winkel

Title: “The Methodological, Epistemological, and Legal Challenges of Studying the Platformized Far Right”
Room: Ideate

Abstract

Before the 2020’s those who did data driven research in the humanities were one-eyed-kings, solitary and often autodidact. We sought peers in different departments, universities and of course in online communities. This also meant that there was limited oversight. In 2018, I decided – accompanied by research platform Dataschool – to explore the possibilities and limitations of doing data driven humanities research at my employer Utrecht University, through taking up the case of GAB, a radical far-right social medium. The ‘Gab Project’ resulted in a conflict, a world class dataset, a dissertation, two theses, an internship, and several publications. The atypical nature of the project posed many methodological, epistemological, and legal challenges. It therefore kicked off an institutional learning process about the possibilities, and best practices for research compliant with General Data Protection Regulation (GDPR). We concluded that scholars that use digital tools and data methods for capture and analysis of web platforms must become literate in operating them. Consequently, data-driven research on the far right is naturally interdisciplinary and therefore cooperative and adherent to the principles of open science.

13:30 – 14:30 Panel discussion

Panel members: Carola Veenstra-Jager, Paddy Leerssen, Roel Lutkenhaus, Giovanni Spitale
Moderator: Efe Sözeri
Room: Ideate

During this panel discussion, experts from the ethics, legal, and privacy fields discuss use cases of social media research.

15:00 – 16:00 Parallel sessions

Session 1: Giovanni Spitale

Title: “Research with Social Media Data: Ethical Issues, Risk Mitigation, and Responsible Practices”
Room: Ideate

Abstract

Social media data offers unprecedented opportunities for understanding human behavior, public discourse, and health-related phenomena, but it also raises complex ethical, legal, and privacy challenges. This lecture explores the distinctive issues involved in working with social media data, including informed consent in public or semi-public spaces, data anonymization

and re-identification risks, platform governance, and cross-border data sharing. Participants will gain practical insights into identifying ethical risks across the research lifecycle, applying mitigation strategies, and implementing responsible data management practices when using digital trace data. Drawing on real-world cases, the session bridges theory and practice to support ethically robust research involving social media and online communities.

Session 2: Karel Kroeze

Title: “Social media research in practice: a hands-on introduction"
Room: Inform

Abstract

As data scientists at BMS Lab and the Behavioural Data Science incubator at the faculty of Behavioural, Management, and Social sciences, we regularly assist our colleagues and students in using social media data in their research, theses, or education. While social media data can be a rich source of data for a wide range of topics – using it effectively also comes with a unique set of practical challenges.

This hands-on workshop gives a short introduction to the practical aspects of using social media data in research. Participants will learn the fundamentals of data science using social media, performing all steps of a typical data science project:

• Identifying sources: Identifying strengths and weaknesses of different social media and picking the right platform for your project.
• Data acquisition: Gaining access and programmatically downloading datasets using Python or R.
• Data exploration: Understanding data structures and identifying research possibilities within social media data.
• Analysis: Performing basic sentiment analysis and exploring individual research questions.

Using a simplified research question as a case study, we will highlight and discuss some of the main decisions and challenges data scientists encounter doing social media research.

Example code and documentation will be available for both python and R so that participants of all experience levels can follow along. That said, participants are encouraged to shape the example according to their interests, experiment with their own ideas, and expand on the provided materials. 

Note: Participants must bring their own laptop and are assumed to have a basic working knowledge of either R or Python. 

Session 3: Danielle McCool

Title: “Data Donation as a Method for Social Media Research: Workflow, Privacy Considerations, and Practical Challenges”
Room: Learn-X

Abstract

Social media researchers increasingly face a difficult trilemma: platform APIs are being shut down, web scraping raises legal and ethical concerns, and the content in freely available datasets rarely aligns with what researchers want to know. Data donation – asking participants to share their own Data Download Packages (DDPs) – offers a promising alternative grounded in GDPR’s right of data access and genuine participant consent. But in its simplest form, data donation is not inherently privacy-preserving: for example, a participant who agrees to share their full Instagram download package would be handing over potentially years of private messages, location data, pictures, or videos, with privacy protection relying entirely on researcher discretion.

In 2022, we developed an approach to address this issue. Rather than asking a participant to donate their whole package, all data extraction and filtering happens locally on the participant’s own device. This allows participants to inspect exactly what will be shared before they consent to donate it, which makes informed consent actually possible rather than just a formality. But each step towards stronger privacy protection introduced its own layer of complexity. In this workshop, we share the full journey: the challenges of obtaining ethical approval, complications related to working with external software development companies, and the ongoing balancing act between what researchers need, what platforms allow, and what participants can reasonably be asked to do.

Participants will experience the data donation workflow firsthand by requesting and exploring their own DDP and going through the local processing and consent flow, making the privacy stakes and methodological choices concrete and personal. We also discuss what researchers need to consider to use data donation responsibly in their own work, from selecting the right platform and constructs to communicating privacy guarantees clearly and navigating a rapidly changing platform landscape.

Please bring a laptop. This workshop begins and ends from the perspective of a research participant. We will request a Data Download Package at the start and explore its contents together at the end. No other preparation is necessary ahead of time.

16:00 – 16:30 Keynote speaker

Jessica Taylor Piotrowski

Title: “Measuring Accountability”
Room: Ideate

Abstract

Europe has set an ambitious goal: hold large online platforms accountable for the risks they create. Under the Digital Services Act (DSA), platforms must assess and mitigate “systemic risks” - such as the amplification of disinformation, harmful content affecting minors, or design features that encourage excessive use. Researchers now have new rights to request platform data to study these risks independently. This is a major step forward. But still there is a practical challenge: how do we determine whether platforms have reduced systemic risk (or not)?

Systemic risks do not sit in a single post or a viral moment. It is about patterns, for example, how recommender systems amplify certain content, how interface features like autoplay or infinite scroll affect behavior, or how exposure accumulates over time. Testing whether companies have taken “appropriate and proportionate measures” (as required under Article 28) requires connecting platform design to measurable real-world outcomes. That requires access to sensitive data, secure environments to analyze it, and clear standards for what counts as evidence. Right now, these conditions are fragile and uneven.

In this keynote, I reflect on what it takes to test systemic risk. Drawing on my experiences, I examine the gap between regulatory promise and research reality. I discuss why data access alone is not enough, why secure and standardized research infrastructures are essential, and why data stewards, infrastructure builders, and social scientists must work together if the DSA is to succeed.

The core argument is simple: platform accountability depends on measurement, and measurement depends on infrastructure. If we want platform governance to be evidence-based rather than symbolic, we must invest not only in regulation, but in the scientific capacity to test it

Keynote speakers

Jessica Taylor Piotrowski

Jessica Taylor Piotrowski, Ph.D., is a Professor in the Amsterdam School of Communication Research (ASCoR), University of Amsterdam, where she holds the Chair of Communication in the Digital Society. She also serves as Director of the Center for Research on Children, Adolescents, and the Media (CcaM), Director of the Graduate School of Communication, Co-Director of the Faculty Teaching & Learning Centre, and Faculty Lead for the national “Human Factor in New Technologies” sector programme.  Her research focuses on how individual and socio-cultural differences shape young people’s selection, use and processing of digital media—and how those processes contribute to cognitive, emotional and social outcomes. A central theme of her work is digital competence: the combination of digital skills and knowledge that enables young people to thrive in a digital society.  Beyond her scholarly work, Piotrowski is committed to translation of research into practice: she regularly engages with stakeholders such as the European Commission, Google, Meta and YouTube, serves on international boards and committees, and champions open science and accessible research. In her leadership and teaching roles she emphasizes reflexivity and community and drives curriculum innovation to meet the changing demands of students and society. Outside of work, she enjoys running (often for charity), dark chocolate and espresso, and is always looking for sunshine in Amsterdam.

Tim de Winkel

Tim de Winkel is dr. in media studies with an expertise in radical platform technology and digital humanities methodology. His dissertation was on far-right fringe platforms. He is currently employed by Erasmus university and the Utrecht University of applied sciences, where he works on several projects regarding 'datafied communication and surveillance by Dutch public institutions and local government' and supervises a project on 'activist and government encounters'. Tim is a Union rep and parlement member for the AOB and FNV, an academic activist for 0.7 and VSMP. He also discovered that the real author of the Dutch national anthem is Petrus Datheen.

Title: The Methodological, Epistemological, and Legal Challenges of Studying the Platformized Far Right