vai al contenuto della pagina vai al menu di navigazione
 

Statistics seminar 2018: Non-parametric micro Statistical Matching: methodological advances and two frontiers application

15/10/2018 dalle 14:00 alle 16:00

Dove Dipartimento di Scienze Statistiche - v.le Berti Pichat - Aula VII

Aggiungi l'evento al calendario

Relatore
Riccardo D’Alberto

Abstract
Non-parametric micro Statistical Matching imputation techniques, the so-called “hot deck” methods, are commonly used to integrate data and overcome several issues, first of all, elementary data shortage.
These techniques received an increasing attention in the last decade due to their flexible use and the computational advantages they offer w.r.t. the parametric methods, but also due to the possibility they offer to avoid model misspecification bias. Further developments and theoretical advances concern the study of new combinations of the existing techniques and not-default distance functions, the assessment of the imputation goodness, the analysis of the commonly accepted and applied “hot deck practical prescriptions”
This seminar will provide insights on non-parametric micro Statistical Matching, discussing the aforementioned issues and methodological challenges. It will also cover two application examples, one related to farm data (the integration of different farm data sources, both official statistics and a project survey) and the other concerning a benefit value transfer exercise.

Organizzazione
Daniela Cocchi