ESRA 2027 sessions by theme
Back to Overview of Sessions
What Can we Learn from Web Panel Metadata |
| Coordinator 1 | Dr Meredith Massey (National Center for Health Statistics) |
Session Details
Session Abstract
Web-based probability and nonprobability panels are increasingly used to collect timely, cost-effective survey data. While researchers typically focus on respondents’ answers, web panels also generate a wealth of metadata and paradata that can provide insight into how respondents interact with surveys and how the data collection process may affect data quality. These data may include response times, device type, breakoffs, survey entry and completion patterns, recruitment and participation histories, mode of completion, panel tenure, and other information about respondents’ interactions with the survey and panel.
This session invites research examining how web panel metadata and paradata can be used to better understand survey participation, respondent behavior, measurement, and data quality. Of particular interest are studies that move beyond describing metadata to evaluate what these measures reveal about potential sources of Total Survey Error. Topics may include relationships between response behavior and data quality; panel conditioning and respondent experience; satisficing and speeding; breakoffs and nonresponse; device and mode effects; patterns of repeated participation; and the use of metadata to identify respondents or survey questions associated with increased measurement error.
The session also welcomes research exploring how metadata can be combined with survey responses, methodological experiments, qualitative research, or external benchmarks to better understand the mechanisms underlying observed response patterns. Papers addressing the practical and ethical challenges of using web panel metadata are also encouraged.
Together, these studies can demonstrate how information generated during web survey administration can be transformed from an operational byproduct into a methodological resource for understanding respondent behavior, evaluating survey quality, and improving the design of web-based data collections.
