SHP_IV Pilot Waves 1-2

In preparation of the next refreshment sample of the Swiss Household Panel (the SHP_IV, to start in 2020), the SHP ran a two-wave pilot study in 2018 and 2019 to test alternative modes of interviewing. Advances in information and communication technol-ogies and the increasing use thereof, suggest that web-based data collection offers a particularly promising alternative to telephone in the Swiss context. However, differences between CATI and web are large in terms of interviewer presence, interview pace, and the channels of communication used (oral or visual) (Couper 2011), meaning the compa-rability between the modes should be evaluated. Also, the SHP is considering recruiting a refreshment sample in a different mode than the one used until now, which for large-scale household panel surveys has not yet been tested. It is often assumed that web mode would be inappropriate for panel recruitment purposes (Tourangeau 2017) because of its typically lower response rates, but it is un-clear whether this argument still holds in the context of a panel recruited by telephone. To test the ways in which offering web as an alternative mode affects response rates, sample composition and measurement, the SHP mounted a two-wave pilot study, incor-porating a mixed mode experiment in 2018-2019. The aim was to compare the standard SHP CATI-based fieldwork (and recruitment) strategy with two online alternatives: a mixed mode (telephone for the HRP interview plus web for individual household mem-bers) group and a web-only group (see Voorpostel et al., in press). Now the pilot study is completed, the data will be made available to SHP data users. These data are especially suited to answer methodological research questions related to interview mode in household panels. As the study is based on a stratified random sam-ple and uses the complete SHP questionnaires, the data can also be used for substan-tive analysis. It is important to note, however, that the weights provided with the data do not weigh for the mode of data collection and the sample of the pilot cannot be easily combined with the main samples of the SHP.