Abstract

Smart surveys employ features of smart devices. Keeping respondents at the heart of data collection, they form a bridge between survey and big data. The incentive to go smart is especially strong for surveys that are considered burdensome, non-central to respondents and/or contain topics for which questions form weak proxies for the concepts of interest. In smart survey data collection AI and machine learning methodology play a prominent role in the transformation of ‘smart’ data to statistics. Smart surveys, thus, follow the potential of modern technology that respondents may appreciate or even expect.
Smart Survey Implementation (SSI) is a project funded by Eurostat and conducted by a consortium of seven national statistical institutes, three universities and one university spin-off. It develops, evaluates and implements smart surveys at all design levels at the hand of three topical applications: household expenditures, time us and household energy usage. The three case studies form a stepping stone to generic procedures and guidelines for surveys that include smart features.
SSI aims to involve and engage citizens in the policy design and evaluation which potentially contribute to the gain of trust and participation of the citizen. This can only be achieved by diving into the respondent perceptions within realistic and legitimate smart survey settings. As part of the project, a cross-national survey on smart survey perceptions has been conducted. The survey offers an empirically supported understanding of how citizens feel about surveys with smart features, including how well they understand what is being measured and what they consent to.
In the paper, the place of smart surveys in future survey solutions is discussed.