ESRA 2027 sessions by theme
Back to Overview of Sessions
Detecting Careless Responding in Online Panel Surveys |
| Coordinator 1 | Mrs Sarah Forrest (National Center for Health Statistics, Centers for Disease Control and Prevention) |
| Coordinator 2 | Mrs Katherine Irimata (National Center for Health Statistics, Centers for Disease Control and Prevention) |
Session Details
Online survey panels promise speed, flexibility, and access to hard-to-reach populations, yet they raise a pressing question: Are respondents actually providing thoughtful answers? Panel surveys now dominate online survey data collection. They range from probability-based panels recruited through established sampling frames to opt-in panels that recruit volunteers. Across both types, key questions remain: How does response quality differ by panel source? How do respondents engage with survey questions? When do panelists cut corners to reduce cognitive effort? Panel features such as the absence of an interviewer, frequent survey invitations, and incentives can contribute to low-effort responding, which affects measurement quality, survey estimates, and Total Survey Error. As the use of online survey panels continues to grow, understanding, detecting, and reducing low-quality responding has become central to survey data quality.
Researchers have used behavioral, psychometric, and attention-based indicators to identify low-quality responding in the U.S. They have also examined response quality and cross-national comparability in multinational probability surveys in Europe. Yet the rich paradata available in panel surveys provide new opportunities to understand response patterns and push this work further.
This session invites papers on satisficing, careless and insufficient effort responding, and response quality in online panel surveys. We welcome contributions related to:
- Measurement and validation of response-quality indicators
- Statistical and machine-learning approaches to identifying low-quality cases
- Comparisons across panel types (such as probability and opt-in), modes, and providers
- Relationships between response behavior and demographic and other respondent characteristics
We encourage cross-panel and cross-national comparisons, as well as methodological approaches to preventing, detecting, or reducing low-quality responding, and frameworks for deciding when to flag or exclude low-effort cases.
