Reducing selection bias in non-probability sample by Small Area Estimation.

Nowadays, the availability of a huge amount of data produced by a wide range of new technologies is increasing. However, data obtainable from these sources are often the result of a non-probability sampling process. We propose a method to reduce the selection bias associated with the big data in the context of Small Area Estimation. Our approach is based on data integration and it combines a big data sample and a probability sample. Real data examples are considered in the context of Italian enterprises sensitiveness towards Sustainable Development Goals and ecommerce.