Current Innovations in Probability-based Household Internet Panel Research
Thursday, 4 March and Friday, 5 March 2021
- Transparency and best practices in probability-based panel recruitment and retention practices;
- COVID pandemic and its effects;
- Survey methodology and practice;
- Data management and linkages;
- Developments in analysis techniques;
- Applications of research;
- Combination of wearable devices data with self-reports;
- Naturally occurring data; and
- Machine learning.
Agenda and Presentations
Please click here for the two-day agenda.
For a full list of speakers, please click here.
Registration is now closed. Please check later in the year for new on CIPHER 2022.
The call for papers is now closed.
CIPHER 2021 was held virtually on Zoom. .
What Internet panel surveys can learn from the design of smartphone-app studies. - Dr. Peter Lugtig
Several studies have explored using smartphone-apps as a way to collect social and behavioral data. Some notable successful studies have in recent years been conducted in the context of travel behavior, time use, household consumption, or 'in-the-moment' attitudinal measures about a variety of subjects.
One characteristic of smartphone studies is that they follow participants over a period of time to measure (short-term) changes. This is a goal that they have in common with panel studies, that typically re-interview the same respondents every year or every few months. Apart from overlap in the goal in these types of studies, smartphone-app studies and panel studies have similar problems in terms of dropout, and panel maintenance.
There are however many differences between panel studies and smartphone studies, that partly stem from the primary device that is being used to collect data. Panel studies have either been designed for face-to-face completion (often re-interviewing respondents every year), or for PC/laptop completion (often re-interviewing respondents monthly or bi-monthly. Smartphone studies usually re-interview respondents daily, or even take continuous measurements, but then last perhaps a few weeks or months in total.
In his keynote, Dr. Lugtig will discuss existing design differences between panel studies and smartphone-app studies using examples from several existing studies. He will discuss the relative strengths of smartphone-apps and panel studies, and their relative weaknesses in the context of the Total Survey Error framework, and will discuss with an overview of what each type of study can learn from each other. To conclude, Dr. Lugtig will outline some possible designs for 'hybrid' studies that use the smartphone to study change over longer periods of time, while including 'measurement bursts' that can be employed to better study life events or short-term change.
This conference has been made possible through funding from a grant partnership between the National Institute on Aging and the Social Security Administration.
Closed Captioning available upon request.