Interventions to Socio-Technical Problems: Data Literacy and the Case of Online Hate Speech

Abstract

Socio-Technical Problems and the Need for Data Literacy In times of accelerating technological and societal change, the ability to understand and critically engage with data has become a key prerequisite for meaningful participation in research, public discourse, and policy-making. Sociological research is increasingly expected to provide insight into complex issues such as misinformation, algorithmic discrimination, and online hate speech. These challenges are deeply entwined with digital infrastructures and data practices and require not only analytical tools but also reflective, theory-informed approaches. Following Beck’s concept of the risk society (1986), such problems are not merely side effects of digitalization, but emerge directly from modernization itself and demand new forms of knowledge to be addressed. Giddens’ notion of reflexive modernity (1990) emphasizes the need for continuous, systematic observation and analysis to adapt to social change. However, as Bourdieu (1984) argues, access to the resources and competencies required to engage in knowledge production remains unequally distributed, particularly with regard to data-related skills. The digital transformation of society has further intensified these dynamics. New types of data—such as social media traces, sensor data, or digital archives—have greatly expanded the empirical base of sociological research (Macy, 2015). At the same time, analytical methods from fields like computer science, natural sciences, and the life sciences are increasingly entering sociological research practice (Tindall et al., 2022)—among them machine learning (Borch & Pardo-Guerra, 2023) and text mining (Macanovic, 2022). These developments demand not only technical expertise but also the ability to critically assess data practices in light of ethical, legal, and social aspects (ELSA). Against this backdrop, data literacy—defined as the ability to collect, manage, evaluate, and apply data in a critical and responsible manner (Schüller, 2023)—becomes a foundational competence in the social sciences. Closing the Gap: Building Data Literacy through Institutional Infrastructure While data literacy is increasingly recognized as essential, it remains underdeveloped in practice, particularly in the social sciences. Many researchers lack access to adequate training opportunities, infrastructure, and methodological support to meet the demands of the changing data landscape. This results in a disconnect between the potential of sociological engagement with socio-technical problems and the resources available to do so. To address these challenges, the interdisciplinary data competence center “DataNord” has been established for the Bremen region. It is part of the U Bremen Research Alliance—a network of the University of Bremen and twelve non-university research institutes—and has been developed in cooperation with additional regional partners. DataNord strengthens researchers’ data literacy across key areas such as data management, data science, critical thinking, and ethical, legal, and social aspects, while also promoting the transfer of this knowledge to society. The following sections present our approach to answering these questions through integrated measures of training, consultation, networking, and knowledge transfer from research projects. The interdisciplinary project DataNord addresses this gap at a structural level by asking: 1. How can we build sustainable, inclusive infrastructures with demand-driven services that effectively strengthen data literacy across all disciplines, with one particular focus on the social sciences? Exemplifying this broader effort, the research project “We the Social Media” based within the DataNord Research Academy, explores how interdisciplinary research and hands-on data work can help address complex socio-technical issues such as online hate speech: 2. How can data literacy serve as a key competency in empowering research to understand and address the socio-technical problem of online hate speech? An Infrastructure to Foster Data Literacy: The DataNord Project DataNord is built around four interlinked components: learning, consulting, networking, and competence development through hands-on research. As part of the learning activities under the DataNord umbrella, the Data Train program offers foundational courses on data literacy, while the University of Bremen’s Data Science Center (DSC) (Steinmann et al. 2023)—with an interdisciplinary team of five data scientists—supplements this by designing in-depth and discipline-specific training formats that introduce social scientists to key concepts and tools of data management, computational methods, and critical interpretations. Through tailored consulting offers, individual researchers and projects receive individual support for planning, executing, and reflecting on their data strategies. Networking activities bring together actors from research, infrastructure, policy-making, and the public to foster interdisciplinary collaboration and sustainable knowledge exchange. The DataNord Research Academy extends DataNord’s service portfolio. It facilitates collaboration between methodological and domain experts across five interdisciplinary research projects that address key challenges in data-intensive research. By fostering advanced data skills and embedding them into the research process, the Research Academy aims to strengthen data literacy and advance data culture in the scientific community. Participatory Research as Intervention: The Case of “We The Social Media” The overarching goals of the DataNord initiative can be concretely illustrated by introducing one of its associated research projects: “We The Social Media”. The Research Academy anchors data skills through hands-on research, fostering advanced expertise. This project illustrates how research can engage with pressing societal issues like hate speech on digital platforms. Hate speech on social media is a well-documented issue (Castellanos, 2023). It not only disrupts public discourse (Stępień-Załucka, 2024) but also incites violence against marginalized communities (Williams, 2020) and undermines social cohesion (Soral, 2018). However, despite its severe societal consequences, current strategies for managing hate speech remain inadequate. Third-party audits of commercial content moderation APIs have shown that these systems perform poorly, particularly in detecting implicit hate speech (Hartmann, 2025). Alarmingly, their performance is weakest when it comes to hate speech targeting LGBTQ individuals and ableist speech (Hartmann, 2025). Compounding this issue is the trend of reduced content moderation efforts by major platforms (Meta, 2025). To address this issue, “We The Social Media” employs user-centered, innovative research methods to better understand the expectations and experiences of social media users, especially those most affected by hate speech. Building on these insights, the project develops and evaluates a participatory design intervention. At the core of the project is a Delphi study involving members of social movements. This sample choice reflects the unique position they occupy: they rely heavily on social media for outreach and advocacy (Hwang & Kim, 2015; Momeni, 2017), while also being particularly targeted by hate speech (Castillo-Esparcia, 2023). The Delphi method, which facilitates anonymous structured communication among a panel of experts through multiple rounds of questioning and aggregated feedback (Cuhls, 2023), was used to gather insights from the activists. The study aimed to identify how they define effective moderation, what control mechanisms they envision, and how they would utilize them. Findings from this study are informing the design and evaluation of a moderation tool tailored to the needs and expectations of these users. This iterative, participatory process not only generates data-driven insights into the social problem of online hate speech but also evaluates the effectiveness of potential interventions, offering guidance for policymakers, platform designers, and civic tech developers. Moreover, the project contributes to the goals of the Research Academy, which anchors data skills through hands-on research, fostering advanced expertise. It enables interdisciplinary exploration of complex socio-technical problems by combining sociological and computational approaches. In doing so, it generates novel data and introduces innovative methodological approaches that support both academic research and public discourse on digital communication and platform governance. Conclusion: Data Literacy as a Sociological Tool for Societal Intervention Through this dual approach—combining infrastructure development, training, and participatory research—DataNord demonstrates how data literacy can be mobilized as a sociological tool for empirical intervention in socio-technical problems. The case of “We The Social Media” shows how research can not only produce actionable knowledge on platform governance but also inform capacity-building efforts and infrastructure design. By closing the loop between research and training, DataNord helps enable researchers and affected communities alike to navigate and shape digital society. Literature Beck, U. (1986). Risikogesellschaft. Auf dem Weg in eine andere Moderne. Frankfurt: Suhrkamp.Borch, C., & Pardo-Guerra, J. P. (2023). Sociology in the Age of Artificial Intelligence. Theory and Society, 52, 231–259. Bourdieu, P. (1984). Science of Science and Reflexivity. Cambridge: Polity Press.Castaño-Pulgarín, S. A., Suárez-Betancur, N., Vega, L. M. T., & López, H. M. H. (2021). Internet, social media and online hate speech. Systematic review. Aggression and Violent Behavior, 58, 101608. Castellanos, M., Wettstein, A., Wachs, S., Kansok-Dusche, J., Ballaschk, C., Krause, N., & Bilz, L. (2023). Hate speech in adolescents: A binational study on prevalence and demographic differences. Frontiers in Education, 8. Castillo Esparcia, A., Caro Castaño, L., & Almansa-Martínez, A. (2023). Evolution of digital activism on social media: Opportunities and challenges. Profesional de La Información, 32(3). Cuhls, K. (2023). The Delphi Method: An Introduction. In M. Niederberger & O. Renn (Hrsg.), Delphi Methods In The Social And Health Sciences: Concepts, applications and case studies (S. 3–27). Springer Fachmedien. Giddens, A. (1990). The Consequences of Modernity. Cambridge: Polity Press.Hartmann, D., Oueslati, A., Staufer, D., Pohlmann, L., Munzert, S., & Heuer, H. (2025). Lost in Moderation: How Commercial Content Moderation APIs Over-and Under-Moderate Group-Targeted Hate Speech and Linguistic Variations. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI '25). Association for Computing Machinery, New York, NY, USA, Article 175, 1–26. Hwang, H., & Kim, K. (2015). Social media as a tool for social movements: The effect of social media use and social capital on intention to participate in social movements. International Journal of Consumer Studies, 39(5), 478–488. Macanovic, A. (2022). Text Analysis for Social Research. Sociological Science, 9, 25–47. Macy, M. (2015). Digital Footprints and Social Research. Science, 347(6221), 2.Meta. (2025, January). More speech and fewer mistakes. Meta. https://about.fb.com/news/2025/01/ meta-more-speech-fewer-mistakes/Momeni, M. (2017). Social media and political participation. New Media & Society, 19(12), 2094–2100. Soral, W., Bilewicz, M., & Winiewski, M. (2018). Exposure to hate speech increases prejudice through desensitization. Aggressive behavior, 44(2), 136-146.Steinmann, L., Thöricht, H., Zänkert, S., & Drechsler, R. (2023). Das Data Science Center an der Universität Bremen: Interdisziplinärer Knotenpunkt und Service-Infrastruktur für die datenintensive Forschung. In E-Science-Tage 2023 - Empower Your Research – Preserve Your Data, edited by Vincent Heuveline, and Nina Bisheh, 246-252. Heidelberg: heiBOOKS.Stępień-Załucka, B. (2024). Freedom of Expression and Hate Speech Regarding the Activities of Public Officials in Political Debate. Ius Novum. Tindall, D. B., Cormier, J., & Dawkins, B. (2022). Computational methods in environmental sociology. Environmental Sociology, 8(4), 314–327