ESRA 2027 sessions by theme

Back to Overview of Sessions

Session Details

The collection of survey data is only one step on the path to the data’s use. Survey data can be complex (e.g., including thousands of variables, split across multiple data files, adhering to inconsistent metadata standards, and so on), hindering use, and linking with external datasets (e.g., geographic data) can greatly expand research utility. Large language models (LLMs) and other generative artificial intelligence (GenAI) systems offer a massive opportunity to increase the use and utility of survey datasets by (1) lowering the costs of developing resources that making data easier to use and explore and (2) enabling the creation and linking of novel data sources into the survey data.

This session will invite submissions of any project designing to increase the use of existing survey data, with the purpose of sparking interest and ideas for increasing the use of survey data. Examples could include: the creation of natural language search tools to explore survey metadata; the development of code generator tools allowing data users to quickly create analytical code for a dataset; the use of LLMs to produce novel datasets, such classification of articles in local newspapers, that could be linked with survey data to expand research possibilities (e.g., on the effect of local high-profile crime upon public behaviour).