ESRA 2027 sessions by theme
Back to Overview of Sessions
Merging Traditional and Novel Data Sources to Produce New Insights |
| Coordinator 1 | Dr Christina Pao (University of South Carolina) |
| Coordinator 2 | Dr Francesco Rampazzo (University of Manchester) |
| Coordinator 3 | Dr Aliakbar Akbaritabar (Max Planck Institute for demographic Research) |
Session Details
Traditional data sources, such as national or international surveys and administrative data, have long been the cornerstones of the quantitative social sciences. The advent of novel data sources, such as through digital trace data and AI-generated content (e.g., AI “personas” and bots), creates new opportunities and challenges for data analysis.
The following panel aims to collect papers that engage with “mixed data” (akin to “mixed methods” approaches) using traditional and novel data sources. All data sources have their limitations and flaws, and the future of social science may benefit from the triaging of multiple data sources. For instance, we may imagine that new insights can be generated by looking at both administrative data and cell-phone data, human and AI-generated survey responses, stated attitudinal surveys and actual social media behavior—as a few examples of the many possibilities.
We believe that these “mixed data” approaches are going to be a promising pathway forward, in light of the rapid changes that are occurring with the advent of generative AI and changes to national data ecosystems. Many original collections, such as with online surveys, have been increasingly infiltrated by bots, and many countries around the world—such as the U.S. and New Zealand—have been slimming down their Census and other national data ecosystems. Combining data sources may not only be interesting but also necessary as a means of triaging insights about the social world. This panel will be unified, not necessarily in subject or type of data source, but the ability to creatively leverage multiple types of data to answer a research question.
