All time references are in CEST
Automated Preloading and Definition of Panel Samples |
|
| Session Organisers | Ms Stephanie Stuck (SHARE Berlin Institute) Dr Fabio Franzese (SHARE Berlin Institute) Ms Carolina Brändle (SHARE Berlin Institute) |
| Time | Tuesday 18 July, 09:00 - 10:30 |
| Room | |
Panel surveys can use demographic information and responses from previous interviews to enhance their efficacy through methods such as dependent interviewing, where previous answers guide subsequent questioning, and (sub-) sample definition, where specific respondent characteristics inform survey sampling strategies of a follow-up interview.
However, the increasing frequency and flexibility of panel surveys, including simultaneous conducted surveys and multi-mode administration (with or without interviewers), present significant challenges for preloading data. When multiple surveys are conducted simultaneously in the same sample, the question arises as to which information, i.e., which source (respondent self-completion, interviewers, survey institute, sampling/register information), should be trusted and used for subsequent interviews. This issue is closely linked with data cleaning processes, as a high frequency of interviewing may not allow for proper data cleaning before sampling or preloading the next survey wave. Addressing these challenges requires robust software infrastructure and innovative implementation strategies that meet the needs of both researchers and survey institutes.
This session aims to facilitate an exchange of knowledge, experiences, and best practices regarding the technical and conceptual aspects of preloading and defining panel samples. The session invites contributions that encompass several key areas:
• Technological Tools: A demonstration of the software and tools currently employed to preload and manage panel samples efficiently.
• Data Management: Best practices for handling data in the context of preloaded panel samples, ensuring data integrity and usability across survey waves.
• Practical Considerations: Addressing the logistical and operational challenges encountered in implementing preloading strategies across different survey modes.
• Impact on Data Quality: Evaluating how preloading and dependent interviewing impact the accuracy and reliability of survey data.
• Methodological Innovations: Highlighting recent advances and future directions in the field, focusing on innovations that enhance the flexibility and effectiveness of panel surveys.
Keywords: preloading, dependent interviewing
Ms Carolina Brändle (SHARE Berlin Institute)
Dr Fabio Franzese (SHARE Berlin Institute) - Presenting Author
Ms Marlen Paulitti (Centerdata)
Ms Stephanie Stuck (SHARE Berlin Institute)
Mr Iggy van der Wielen (Centerdata)
Ms Sabrina Zuber (SHARE Berlin Institute)
The Survey of Health, Ageing and Retirement in Europe (SHARE) is a comprehensive multi-national panel study that has been conducted for over 20 years, involving approximately 70,000 interviews per wave. Traditionally, data collection occurred every 2 to 3 years using face-to-face surveys. To enhance efficiency and reduce respondent burden, SHARE employs techniques such as preloading and dependent interviewing, particularly in managing household composition data. This approach requires respondents to report only changes in household composition, such as moves or deaths. Household composition data, central to SHARE, have been particularly sensitive to inaccuracies due to the involvement of multiple respondents within households and the inclusion of proxy interviews. Ensuring accurate household composition and maintaining high data quality have required extensive data cleaning and elaborate preload preparation before subsequent survey waves.
As SHARE transitions to more frequent surveys using telephone and web-based methods, new challenges arise, particularly regarding the reduced time available for data cleaning and preload preparation between waves. This shift necessitates the development of new infrastructure to ensure high-quality data collection in a high-frequency, multi-mode environment while remaining fully compliant with GDPR regulations. Additionally, an integrated sampling functionality is planned to enable dynamic sub-sampling during survey operations. In this presentation, we introduce SHARE’s future tools and infrastructure designed to meet these challenges, thereby ensuring high-quality data collection in future panel surveys.