ESRA 2027 sessions by theme

Back to Overview of Sessions

Session Details

Open-ended questions provide respondents with the opportunity to answer in their own words and can capture information that is difficult to obtain through predefined response categories. At the same time, processing open-ended responses is often resource-intensive and poses challenges for data quality, comparability, and subsequent analysis. Recent advances in automated methods offer new opportunities to support the collection, coding, and processing of open-ended survey data.

This session invites contributions examining automated approaches to transforming open-ended survey responses into high-quality structured data, both during and after data collection. Relevant approaches include autocomplete and lookup functions, automated and semi-automated coding, artificial intelligence (AI) methods such as natural language processing, machine learning and large language models, as well as other forms of computer-assisted processing.

We particularly welcome empirical studies, including survey experiments, that investigate their implications for coding accuracy, data quality, respondent burden, response behaviour, processing efficiency, and costs, including both direct survey costs and indirect costs (e.g., environmental costs of AI). Contributions addressing methodological and ethical challenges, including ambiguous or incomplete responses, classification errors, bias, privacy, transparency, reproducibility, data governance, and human–machine collaboration are equally encouraged.

By bringing together research on automated assistance during data collection and automated processing after data collection, the session aims to discuss when and how computational methods can improve the conversion of respondents’ answers into reliable and analytically useful survey data, as well as the methodological trade-offs associated with their use.