Email) (1,2), Bart Buelens, (Email) (3), Rainer Schnell">
12:30 - 13:30
Poster Session
Room: Lunches Space
Implementing Big Data in Official Statistics: Capture-recapture Techniques to Adjust for Underreporting in Transport Surveys Using Sensor Data
Jonas Klingwort, (Email) 1, 2, Bart Buelens, (Email) 3, Rainer Schnell, (Email) 2
1 University of Duisburg-Essen, Duisburg
2 Statistics Netherlands (CBS), Heerlen
3 Vlaamse Instelling voor Technologisch Onderzoek (VITO), Genk
Producing unbiased estimates in official statistics based on survey data becomes more difficult and expensive. Accordingly, research on methods using big data for the production of official statistics is currently increasing. Up to now, big data is rarely used in statistical production due to its unknown data generating process. However, in the long-term, using big data in official statistics is unavoidable. Therefore, instead of using single big data sources, research on combining different probability and non-probability based datasets is a promising approach to use big data in official statistics. More specifically, the different problems of surveys and big data might be minimized if the survey and the sensors measure the same target variable and the resulting micro data can be combined with a unique identifier. Using this principle, we link survey, sensor, and administrative data for transport statistics. Using the linked dataset we apply capture-recapture techniques to validate, estimate and adjust a bias due to underreporting in the target variables of the survey.


Reference:
POST01-013
Session:
Big data analytics (poster)
Presenter/s:
Jonas Klingwort
Presentation type:
Poster presentation
Room:
Lunches Space
Date:
Tuesday, 12 March
Time:
12:30 - 13:30
Session times:
12:30 - 13:30