Computational Opportunities and Challenges in the Measurement of Media Use and Media Effects

This call invites contributions that address core opportunities and challenges at the intersection of computational methods and the empirical study of media use and effects. We welcome theoretical, methodological, and empirical work – especially studies that critically reflect on how computational approaches shape what and how we can know about media exposure and its consequences. 

Submissions may engage with, but are not limited to, the following five areas:

Digital media environments generate a wide array of digital behavioral data – from platform APIs and web tracking to sensor data, app usage, and data donations. These sources offer unprecedented access to patterns of media use and exposure. Yet they also raise important challenges related to visibility, granularity, and interpretability. We invite contributions that explore: 

• The methodological integration of digital behavioral data with qualitative insights, surveys, experiments, or psychophysiological measures
• The design of hybrid systems for tracking and analyzing user behavior
• Establishing how computational methods (e.g., clustering, text-as-data, anomaly detection) can support the integration of digital behavioral data into media use and media effect studies
• The comparative strengths and limitations of different data sources in capturing media use and linking it to potential effects 

Contemporary media environments are shaped by personalization, algorithmic curation, and cross-platform fragmentation. Computational methods allow researchers to track exposure and engagement in granular detail, but they also raise fundamental theoretical questions: What constitutes “use” or an “effect”? How can complex, latent phenomena such as belief change, polarization, or emotional resonance be meaningfully modeled and measured across diverse contexts and time frames? We invite contributions that: 

• Advance or revise theories of media choice, exposure, and effects in light of large-scale digital behavioral data
• Operationalize and model latent constructs (e.g., ideology, emotional tone, framing, text complexity) using computational approaches such as simulations and machine learning
• Link media content to user-level outcomes through automated exposure-effect pipelines, including dynamic and cross-platform analyses
• Critically assess the validity and limitations of computational proxies in capturing complex psychological or sociopolitical effects 

Studies based on voluntary participation and the installation of research software (e.g., apps, browser plugins) often suffer from selection effects, while platform data provide only partial access to media use and exposure. These limitations have direct consequences for the validity and generalizability of media effects research. We seek papers that address: 

• Statistical or causal inference methods for mitigating self-selection and representational bias (e.g., weighting, imputation, sensitivity analysis)
• Methods for identifying and quantifying data distortion, including comparisons with benchmark data
• Approaches integrating qualitative methods (e.g., interviews, diaries, ethnographic observations) to contextualize digital behavioral data, provide deeper insights, and identify blind spots and biases in media use and effects research
• Design strategies to broaden participation and enhance validity and robustness in computational measurement 

The promise of computational media research depends on sustainable and transparent infrastructures. Yet many tools are developed ad hoc, lack documentation, or are inaccessible for reuse – hindering cumulative science. We invite contributions that: 

• Document and share reusable tools for collecting, annotating, or analyzing digital behavioral data on media use and media effects
• Discuss standardization practices for tool development and data annotation
• Reflect on the institutional and technical conditions needed for open, reproducible, and scalable computational research
• Assess the energy and resource sustainability of computational pipelines as well as the ethical and labor dimensions of data work (e.g., conditions in data labeling) 

The analysis of digital behavioral data raises complex ethical and legal questions – particularly when it involves sensitive topics, long-term monitoring, or vulnerable populations. Computational media research must navigate these challenges responsibly and transparently. Relevant topics include:

• Consent, privacy, and data protection in passive and long-term digital behavioral data collection
• Normative frameworks for interpreting and communicating media use and effect findings
• Ethical uncertainties and legal grey areas in cross-platform, multi-method research designs 

Submission Format 

We welcome conceptual, empirical, and methodological contributions. Submissions related to the conference theme may address the suggested areas outlined above but are not limited to them.

Open Panel 

In addition to contributions aligned with this year’s conference theme — Computational Opportunities and Challenges in the Measurement of Media Use and Effects — the conference will host one or more Open Panels for work that lies beyond the central theme. Submissions to the Open Panel should demonstrate clear methodological relevance. Please indicate “Open Panel” in your submission. The same formal criteria for submission and review apply as for presentations within the main conference theme.

Interactive Formats

Submissions are also invited for interactive formats beyond traditional conference presentations. One-hour slots are available for proposals that include a description of the topic, the participants involved, and the planned format or structure. Both discursive formats (e.g., thesis café) and practical formats (e.g., workshop) will be considered. The goal of these interactive formats is to broaden or deepen the conference theme on a different level. Acceptance decisions will be made by the organizers in consultation with the division’s speakers, taking into account the overall program.