PhD Projects @CCL
The PhD projects conducted at the Computational Communication Science Lab cover a broad range of methods and topics related to political communication. We here introduce the projects ongoing at the moment in the order of their starting dates and provide links for the related publications. Concluded PhD projects can be accessed via this link.
Muhammad Noor Fakhruzzaman: Issue-centered framing in Indonesian election news and the influence of media ownership
Media ownership in Indonesia is concentrated among elites with governmental ties. These owners may influence the writing direction of election reporting in Indonesia at the organizational level. While issue-centered reporting is important in informing the public about the policy and the governance of the country, media owners may avoid deeper discussion of issues by using non-substantive frames. This research investigates how Indonesian election news is framed using a mixed-method of computational and qualitative methods. It is a cumulative dissertation, divided into three interconnected studies. The first examines framing characteristics in online Indonesian election news from 2019 to 2024 through qualitative content analysis, and to discover the motivations of journalists in framing preferences in election news. The second study tests IndoBERT, a pre-trained Indonesian language model, to classify the frames and compares it with multilingual models, M-BERT and Google Gemma, then integrates it into a browser extension as a ready-to-use tool. The third study compares framing usage between conglomerate-owned and independent media, exploring the influence of media ownership on election news. The study aims to understand the role of media ownership in news content and explore the possibilities of a low-resourced language model to recognize framing patterns in election news. Finally, to create awareness about the importance of issue-centered journalism for rational political decision making and to strengthen democratic engagement in Indonesia.
Start: March 2025
Supervisor(s): Annie Waldherr, Aytalina Kulichkina
Funding: Ernst Mach Grant – ASEA-UNINET OeAD
Moritz Sedlatschek: You can do it! Social Media, Influencers, and the Perception of Economic Inequality
Economic inequality poses significant risks to democratic stability, yet public concern and awareness remain limited. Despite growing scholarly interest, we still lack a systematic understanding of how social media in particular, contribute to public perceptions of inequality. This dissertation investigates how social media, particularly influencer content and platform algorithms, shapes individuals’ perceptions of economic inequality, integrating theories of the malleability narrative, cultivation, and social learning. The project explores how engagement with influencers’ aspirational content affects users’ beliefs about social mobility and meritocratic beliefs, how influencers and audiences co-construct narratives of success and merit, and how (trans)parasocial relationships with visibly successful figures online complicate our understanding of social segregation and the inference of economic inequality. Ultimately, this research aims to clarify the role of social media in shaping economic worldviews—and, more broadly, to inform debates on (entertainment) media influence, inequality, and the resilience of system justifying ideologies.
Start: October 2024
Supervisor(s): Annie Waldherr, Fabienne Lind
Funding/Projects: CCL
Veronika Ebner: Shaping Climate Conversations: Constructive Communication and Public Engagement on Social Media
Climate change communication has long been dominated by distant threat narratives and fear appeals, yet, mobilizing climate action and behavioral change remains a major issue. Evolving out of the critique that the negativity bias within climate change communication causes unfavorable consequences such as audience disengagement, avoidance or climate anxiety, alternative approaches following positive psychology promote more solutions-oriented climate narratives. While constructive and solutions journalism have gained traction in traditional media contexts, their application in the participatory and multimodal landscape of social platforms remains underexplored. There is some indication that next to news media, other actor types, e.g., climate influencers or activists, adopt a similar shift in their communication, applying more solutions-oriented, hope-inducing narratives. Yet, it is unknown to what extent this is happening and which constructive elements are used by which actors with what influence on user engagement. Therefore, this dissertation aims to address this gap within three interrelated studies: First, by developing a computational, multimodal measurement pipeline to classify key dimensions of constructiveness in climate-related social media posts. Subsequently, I will apply this pipeline to compare actors in their usage of different elements and investigate which sectors are communicated more solutions-oriented. Lastly, I want to analyze audience reactions to constructive content, examining engagement patterns and emotions within comment sections. Theoretically and methodologically, my dissertation aims at advancing conceptualizations and measurements of constructive communication beyond traditional textual news content. Practically, findings will inform climate communicators about effective strategies for engaging social media audiences with constructive, solutions-oriented content.
Start: October 2024
Supervisor(s): Hajo Boomgaarden, Fabienne Lind
Funding/Projects: CIDAPE
Blerta Blakaj: Media Effects on Emigration Aspirations in Transitional Societies: An Analysis with Special Reference to Television and Social Media in Kosova
This dissertation explores the influence of media content, specifically television news programs and social media images, on individuals' migration aspirations. The role of media in emigration has been understudied, and this research aims to address that gap by examining the relationship between media content, media exposure, and attitudes toward emigration. The study focuses on Kosova as a case study due to its unique characteristics, including its visa policy, migration history, internet usage, and media landscape. Using the CoMiD model (Gruber et al., 2020), a linkage approach, and ESM data, the research investigates how negative news about the country of origin and positive images of life abroad interact to shape emigration aspirations. Furthermore, the study explores how attitudes toward emigration may change after visa liberalization, potentially reducing the media effect. By filling the research gap on the impact of media on emigration, this research contributes to understanding the role of media in migration decisions. It also has implications for developing strategies for migration, promoting democratic development, and empowering societies through targeted media literacy and good governance interventions.
Start: March 2023
Supervisor(s): Hajo Boomgaarden
Marvin Stecker: The Mobilisation of Culture - Automated Methods to Investigate Collective Identities in Political Communication
The dissertation investigates the sources and evolution of collective identities in political communication in democracies. It follows two primary research objectives: the investigation and validation of automated text analysis methods to measure group appeals in political communication, and, subsequently, their application to investigate the longitudinal salience, effects, and predictors of collective identity construction in political discourse. I look at collective identities as a binding glue between the demand and supply side of political change: they are rooted in citizens' perceptions and interpretation of structural and economic forces in their daily life, but their political potential is only realised through the activation and mobilisation by political elites. The dissertation will first compare different methods in their application, accuracy, and validity for the automated detection of group appeals in political texts. After this, the dissertation looks at the political elites themselves and their role in constructing collective identities and associations with political parties. The last study will consider the role of intermediaries, namely journalists, in the allocation of social groups to political parties.
Start: Oktober 2022
Supervisor(s): Hajo Boomgaarden, Fabienne Lind
Funding/Project: OPTED, AuthLIB
Apeksha Shetty: Digital Dissemination of Misinformation — How Social Identity Interventions Can Help
Misinformation is a major societal challenge that can undermine trust in democratic institutions and even contribute to real-world violence. While existing interventions for misinformation tend to focus on correcting inaccurate content, research shows that social identity can strongly influence information processing and beliefs. However, few studies have explored how social identity can be leveraged to counteract misinformation. This doctoral project aims to address this gap in the literature by integrating psychological and communication science research to understand how social identity-based interventions can reduce belief in and sharing of misinformation. Specifically, the project will include (1) an experimental study that reduces social identity threat prior to providing corrective information, (2) an experimental study that draws attention to the inconsistency of misinformed beliefs with social identity, and (3) the simulation of an agent-based model to test how to best enhance the potential success of these interventions by targeting specific users in an algorithmically biased information environment. The goal of the project is to generate potential solutions for combating misinformation rooted in social beliefs as well as generate hypotheses regarding the impact of the social environment in enhancing or inhibiting the effectiveness of such interventions.
Start: Oktober 2022
Supervisor(s): David Garcia (external), Annie Waldherr
Funding/Project: EMOMIS
Jula Lühring: Emotional Misinformation Diffusion: Mapping Socio-Affective Dynamics in the Digital Information Ecosystem
Misinformation spreads by maximally exploiting engagement mechanisms of digital platforms, but the explanations of the circumstances under which accuracy comes secondary are limited. Such explanations typically focus on cognition, but people consider false information truthful, especially on social media, because emotional and social dynamics strongly influence their judgment. The attention-based design of platforms reinforces engagement with untrustworthy content in times of an unprecedented spread of misinformation on social media. In my cumulative PhD project, I will therefore investigate the interplay of emotional dynamics and algorithmic recommendation in the diffusion of misinformation on Twitter. Here, misinformation is measured in terms of tweets containing URLs from sources following untrustworthy practices. Understanding misinformation spreading as a symptom of a polarized information ecosystem, I plan to tackle its complexity with a multifaceted approach that (1) describes socio-affective dynamics in the diffusion of information from sources with varying levels of trustworthiness on social media, (2) characterizes users with an untrustworthy information system and the emotional dynamics leading up to real-world actions, in this case, the participation in protests against COVID-19 measures, and (3) explains patterns of socio-affective misinformation spreading on social media in an agent-based simulation model. Using an interdisciplinary and exploratory approach, I hope to contribute to understanding the emotional dynamics in the diffusion of misinformation on social media and generating new ideas for user-oriented platforms.
Start: Oktober 2022
Supervisor(s): Annie Waldherr, David Garcia (external)
Funding/Project: EMOMIS
Paul Balluff: Multilingual Named Entity Recognition with Neural Networks
This dissertation project aims to push the methodological boundaries of multi-lingual automated content analysis for communication science. Its objective is to first, curate cutting-edge technologies to explore new ways for named entity recognition, co-reference resolution, and disambiguation for monolingual and multilingual applications. Second, increase the accessibility of these tools, and third, to provide best practices and recommendations for computational text analysis. This is needed for the repertoire of automated techniques, because named entities lay the foundation for other steps of computer assisted and automated text analysis, such as document and issue linking, quotation extraction or sentiment. State-of-the-art NER methods outperform previous ones, but are challenging to implement by researchers who are not from the area of computer science. Therefore, this dissertation also aims to curate such tools to close the widening gap. Finally, this dissertation will employ case studies to demonstrate the possibilities of high-performing NER tools. The case studies will cover a large time-span and include documents from different genres and languages.
Start: Oktober 2021
Supervisor(s): Hajo Boomgaarden, Annie Waldherr
Funding/Project: OPTED
Ahrabhi Kathirgamalingam: Pushing the limits? Long-term dynamics of racist discursive patterns in mediated discourses
Shifts in public and mediated discourses towards the normalization of racist views are seen as common explanations for the growing success of right-wing populism and widespread racist attitudes and behaviors. While 2015, the year of increased migration movements, is often regarded as a turning point, literature from the 1990s identifies similar racist discourse patterns in media as more recent studies. A broader look at the last 30 years might provide valuable insights into the dynamics of racist discourse patterns and specific events that have shaped possible shifts. Therefore, this dissertation project seeks to provide an unprecedented long-term analysis of racist discourse patterns in different media contexts. By bridging the fields of communication science, political science, contemporary history, and corpus linguistics, this project does justice to the complexity of the subject of racism and ensures continuity of valuable theories and application of innovative methods. Firstly, an instrument to measure racist discourse patterns will be developed and validated. In subsequent studies, the measuring instrument will be employed and other methods of analysis to examine shifts in traditional press media, alternative media and political communication in social media. Findings will contribute to theoretical and methodological advancements of research on the discursive construction of racism. Furthermore, it is dedicated to shed light on the societal phenomenon of racist imaginaries and, thereby, is relevant for society and practice.
Start: Oktober 2021
Supervisor(s): Hajo Boomgaarden
Funding/Project: sowi.docs stipend
Azade Esther Kakavand: The Networked Counterpublics of the Far-Right – Exploring Connections on and Across Social Media
While numerous aspects of far-right communication are already researched, the connections of far-right communities on and across social network sites (SNS) are not investigated yet. With the emergence of social media, among many other communities, the far-right has found new platforms to communicate among their counterpublic. In my cumulative PhD project, I am going to fill a gap in previous research by (1) conducting a systematic review of existing literature on far-right social media presence, (2) comparing the topics that are prevalent in far-right communities on Facebook, Instagram, Twitter, YouTube, and Telegram, (3) analyzing differences in far-right networks on these platforms, and (4) exploring the connections that can be found across these platforms. I focus on the influence of affordances of social media platforms on the topics and networks of the far-right as well as the question of how far-right counterpublics use a combination of platforms to profit from their different affordances.
Start: October 2020
Supervisor(s): Annie Waldherr, Nicola Righetti
Funding/Project: CCL