Relevance & Research Questions: The paper addresses the expanding possibilities and the methodological challenges of online mobile surveys in Italy. It discusses coverage and provides a discussion on main data quality implications of unintended online mobile interviews. Two main research questions: (a) What is the level of coverage across social strata? (b) Is the public opinion industry aware of the methodological implications of unintended mobile interviews? Do they take advantage of paradata on the type of device used, in order to gain some control over this phenomenon? Do they optimize questionnaire and questions for use on mobile devices, in order to prevent drop-offs and to assure homogeneity in measurement across devices?
Methods & data: results from two related studies are presented. The first one addresses ‘structural coverage’, i.e. the diffusion of web-enabled mobile devices across the different segments of the Italian population, and its dynamics. It also compares it with coverage for other ICT-based survey modes (secondary analysis of data from a large-scale repeated survey, carried out face-to-face on a probabilistic sample).
The second study is based on an original survey of Italian market research companies and pollsters. The aim of the latter is to assess to what extent – and how - Italian industry are dealing with unintended online mobile respondents, in terms of both generalization and measurement.
Results: results from the first study show current limitations, but also the rapid expansion, in coverage for online mobile surveys. It compares it with coverage for the main alternative modes: phone and web surveys, not exempt from coverage problems.
Results of the survey on public opinion companies will be available in February 2014, as the survey field is scheduled for January 2014.
Added Value: The paper provides a contribution – limited to the Italian case – to the relatively underdeveloped methodological literature on online mobile surveys. It discusses in details coverage. It also discuss how the public opinion industry is dealing with unintended online mobile respondents in order to assure data quality.