ESRA 2027 sessions by theme

Back to Overview of Sessions

Session Details

Surveys are increasingly being supplemented by passively collected digital traces, such as activity logs from digital platforms, providing a direct window into participants’ digital behavior. Such traces can be utilized for studying media consumption (e.g., news and entertainment content), information seeking (e.g., health), and the adoption of novel technologies (e.g., Large Language Models). Similarly, traces offer manifold opportunities to shed light on web survey participation behavior, such as multitasking, looking up information to knowledge questions, and providing LLM-based answers. Although methodologists have developed approaches for collecting traces, their subsequent processing, enrichment, and analysis is largely unstandardized, offering little guidance and almost no best practices.

For example, researchers increasingly collect activity logs from digital platforms, such as viewing histories from YouTube, Netflix, and TikTok. To draw substantial conclusions, these traces must often be enriched with contextual data (e.g., video content). Similarly, researchers collect participants’ browsing histories, including URLs related to website visits and online searches. These traces must be classified to measure specific types of digital behavior (e.g., information seeking). However, traces are often ambiguous and their large volume renders the application of automated (classification) methods necessary, introducing new error sources.

This session invites contributions that go beyond the mere collection of traces by explicitly focusing on their processing, enrichment, and/or analysis. We particularly welcome studies that derive best practices and empirically informed recommendations for future research. Topics may include (but are not limited to):

- Trace-based operationalization of digital behavior
- Data integration and linkage as well as quality assessments of contextual data
- Automated classification methods (e.g., LLM-based coding)
- Respondent-based annotation of traces in survey contexts
- Qualitative studies on participants’ interpretations of their traces
- Error sources and error adjustments after data collection
- Impact of analytical decisions and reproducibility of computational approaches