Trusted smart surveys have revolutionized data collection by combining traditional survey techniques with modern technological advancements. These surveys intelligently combine self-report questions with smart features collected via sensor-enabled devices such as smartphones, wearables, and other devices, aiming to enhance data quality, reduce burden on participants, and provide more timely and granular data.
The Smart Survey Implementation (SSI) project seeks to develop, implement, and demonstrate the concept of trusted smart surveys, showcasing a complete, end-to-end data collection process. This involves a) engaging citizens as active contributors, b) acquiring, processing and combining data from smart devices, and c) ensuring strong privacy safeguards. The project adopts an organizational structure based on smart survey design levels, focusing on Methodology (WP2), IT Architecture (WP3), Logistics (WP4), and Legal-ethical (WP5). By prioritizing this design approach over an application-based structure, the project aims to develop, test, and evaluate smart services through topical and realistic case studies.
Task 2.4 concerned data integration and issues of mode effects when bringing together data from smart surveys with data from traditional survey methods. We investigated the extent to which data could remain comparable between the smart and non-smart mode, both with respect to measurement and representation. We found that measurement differences were present between smart- and non-smart surveys and could be quite large. The largest contributors were differences both in extent and the type of missingness, as well as differences in how users entered their data. The task proposed discriminating between high-smart cases, where the raw smart data are likely to be very different from the traditional data, and low-smart cases which are similar to web-based surveys.