Download
slide1 n.
Skip this Video
Loading SlideShow in 5 Seconds..
Discussants Felipa Zabala , Orietta Luzi PowerPoint Presentation
Download Presentation
Discussants Felipa Zabala , Orietta Luzi

Discussants Felipa Zabala , Orietta Luzi

139 Vues Download Presentation
Télécharger la présentation

Discussants Felipa Zabala , Orietta Luzi

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. UNECE Worksession on Statistical Data Editing Oslo, 24-26 September 2012 Session topic (iii) – Editing and Imputation in the context of data integration from multiple sources and mixed modes Discussants Felipa Zabala, Orietta Luzi

  2. Session topic (iii) – Editing and Imputation in the context of data integration from multiple sources and mixed modes Integrating data from multiple sources (including administrative data) is becoming progressively more common in official statistics since it allows an NSI to increase its ability to provide more information of good quality while reducing production costs and respondent burden. This can be considered as a “special case” of mixed-mode data collection where information on target variables is captured via different “tools” (e-questionnaires, paper questionnaires, direct access to companies systems, use of administrative data, …)

  3. Session topic (iii) – Editing and Imputation in the context of data integration from multiple sources and mixed modes In these situations, the differences in quality of multiple sources/mixed modes data is a basic issue: the editing strategy needs to be developed to cover data from each source/mode, as well as to ensure the coherence of the integrated sources.

  4. Session topic (iii) – Editing and Imputation in the context of data integration from multiple sources and mixed modes • Covered issues • Issues emerging in the context of data integration from multiple sources and modes, which are to be aware of when developing editing strategies and systems • • The quality of data from various sources and modes, how it compares across sources (and over time) • • Impact on E&I when integrating new types of data sources (including administrative data) into statistical production processes • • Strategies for better alignment of statistical and administrative/ external data purposes, to improve data quality

  5. Session topic (iii) – Editing and Imputation in the context of data integration from multiple sources and mixed modes • Long Presentations • Norway – Micro integration of register-based census data for dwelling and household, Li-Chun Zhang, Coen Hendriks • Statistical methods and approaches to create a complete census data file of linked dwellings and householdswhich meets the need for detailed tabulation and analyses required by the 2011 register-based Pop census in Norway • New Zealand – All answers? Statistics New Zeland’s Integrated Data Infrastructure, Felibel Zabala, Rodney Jer, Jamas Enright and Allyson Seyb • Issues and solutions relating to the development of the Statistics NZ’s Integrated Data Infrastructure covering different statistical areas and allowing for complex multivariate analysis of social and economic phenomena

  6. Session topic (iii) – Editing and Imputation in the context of data integration from multiple sources and mixed modes • Short Presentations • United Kingdom – Editing challenges for new data collection methods, Rachel Skentelbery Rachel Skentelbery, Carys Davies • France – Methodological Questions Raised by the Combined Use of Administrative and Survey Data for French Structural Business Statistics, Philippe Brion • Hungary – Studying the Options of Substituting a Regular Statistical Survey with Administrative Data, GergelyHorváth, ZoltánCsereháti • Abu-Dhabi – Evolving data processing in the statistics centre, DragicaSarich an d Maitha Al Junaibi • United Kingdom – Editing and imputing VAT for the Purpose of Producing Mixed-Source Turnover Estimates, Hannah Finselbach (presented by Claire Dobbins) • 10:40 Coffee Break

  7. Session topic (iii) – Editing and Imputation in the context of data integration from multiple sources and mixed modes • Short Presentations • Netherlands – Imputing missing values when using administrative data for short-term enterprise statistics, Pieter Vlag • Italy –Improvement of the Timeliness of the Italian Business Register via Imputation of Missing information in Administrative data, D. Di Cecco, D. Filipponi