Advances in Longitudinal Survey Methodology

9Mixing Modes in Household Panel Surveys: Recent Developments and New Findings

Marieke Voorpostel¹, Oliver Lipps¹, and Caroline Roberts²

¹Surveys Unit, FORS, Lausanne, Switzerland

²Institute of Social Sciences, University of Lausanne, Lausanne, Switzerland

9.1 Introduction

Survey researchers have always adapted their methods in response to the latest societal and technological developments (Groves 2011), but the growth of alternative, lower‐cost modes of data collection provided by the internet has resulted in a sea change in how surveys are carried out. Web‐based surveys allow data to be gathered from (more) respondents more quickly and cheaply than is feasible with interviewer‐administered surveys, but they need to be used in combination with other methods to ensure adequate representation. While mixing web with other modes of data collection may be effective at reducing fieldwork costs and selection errors, it can create a number of complications for researchers conducting the survey (in terms of questionnaire development and fieldwork management) and analysing the survey data (because data gathered in different modes may not be comparable) (De Leeuw 2005). These complications are particularly challenging in the context of longitudinal surveys of households, and many large‐scale academic panel studies have been reluctant to move to mixed mode designs. Yet recent years have seen a growing interest in the potential benefits to be gained from mixed‐mode panels, and a growing ...