Big Data in official statistics
Professor: Jan van den Brakel (Methodology Department, Statistics Netherlands, and Department of Quantitative Economics, Maastricht University)
Idioma: Inglês
Data: 22 de maio de 2019
Horário: 14:00 às 17:30
Carga horária: 3 horas
Local: Auditório – Sala 306
Vagas: 50
Inscrições: de 8 a 16 de maio, na Secretaria da ENCE ou por e-mail para
O endereço de e-mail address está sendo protegido de spambots. Você precisa ativar o JavaScript enabled para vê-lo.
.
Pré-requisitos: Conhecimentos de modelagem estatística e métodos para pesquisas e levantamentos.
Conteúdo
National statistical institutes are under increasing pressure to reduce administration costs and response burden for the production of official statistics. This could potentially be accomplished by using large data sets - so called big data. However, there are problems that must be addressed when using such data sources for the production of official statistics.
In these sessions, two different research lines will be presented on how big data sources can be used in the production of official statistics. They will be illustrated with research results from projects conducted at Statistics Netherlands.
The first approach combines big data sources with sample data in a model-based inference approach. This implies the use of big-data sources as covariates in models for small area estimation, for example in an area level model where cross-sectional correlation between areas are taken into account to improve the effective sample size of the domains.
The second approach employs big data as a primary data source for the compilation of official statistics. This can be considered if a big data source covers the intended target population and does not suffer too much from under- and over-coverage, e.g. the use of satellite and areal images for deriving statistical information on land use. In most cases, however, adjustments for selection bias are required.