Panel sample composition development | Munich Center for the Economics of Aging - MEA
01.08.2016 - 30.09.2020 / SHARE - Research

Panel sample composition development

In this project, we study the development of the panel samples in SHARE since the first wave, using an R-indicator. Normally, an R-indicator is calculated as the deviation of a net sample from a representative gross sample or target population, on a number of standard auxiliary variables. For SHARE, we use it in a different way: Instead of taking a representative gross sample or population values as reference, we use the SHARE first wave net sample, after recruitment, as the standard and calculate the deviation of the remaining sample in each wave from that first wave sample. This allows us to include many more variables than only the simple demographics that are available for gross samples. The aim is to give a descriptive statistic of our panel sample selectivity over time and to estimate how our substantive research variables change over time as a result of change in the panel’s composition over time.


Sabine Friedel, M.A.

Dr. Annette Scherpenzeel
Using the R-indicator to study attrition bias in a probability based Web panel

Friedel, Sabine; Birkenbach, Tim (2020): "Evolution of the Initially Recruited SHARE Panel Sample Over the First Six Waves". In: Journal of Official Statistics, 36, 3, S. 507-527.