English title: Assessing representativeness of a probability-based online panel in Germany

Author(s): Bella Struminskaya - Lars Kaczmirek - Ines Schaurer - Wolfgang Bandilla -

Language: English

Type: Book chapter

Year: 2014

Abstract

The growth of online survey research has led to an increased demand for probability-based online panels. Several of such panels are established in Europe and in the USA. As probability-based online panels are being used by scientific institutions to collect data and make inferences about the target population, questions about the quality of such data continue to be raised. In this chapter we assess the quality of an offline-recruited probability-based online panel of Internet users in Germany. First we report the key performance measures of the recruitment and the online surveys. In the second step of the quality assessment we compare our data to other surveys. As benchmarks we use two population surveys: the German General Social Survey (ALLBUS 2010) and the German subsample of the fifth round of the European Social Survey (ESS 2010). Both of these sources contain the information on private Internet usage and thus allow us to compare the estimates from our panel with the estimates calculated for subsamples of Internet users from the reference surveys. Both demographic and attitudinal measures are considered. We assess the feasibility of post-stratification weighting to correct for noncoverage and nonresponse. Additionally, we assess the comparability of the three surveys in modeling social phenomena. This chapter provides insight into the quality of data collected via online panels and discusses the efficiency of probability-based online panels as means of data collection for scientific purposes.

From page no: 61

To page no: 85

Anthology: Online Panel Research: A Data Quality Perspective