8
Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007 IMF Statistics Department (STA). The views expressed in this presentation are those of the author and do not necessarily represent those of the IMF Execuive Board, Management, or staff.

Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

Embed Size (px)

Citation preview

Page 1: Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

Parallel session 3.2: Data Quality & Comparability - System

Kim ZieschangIMF

Fourth International Conference on Agricultural Statistics22 October 2007

IMF Statistics Department (STA). The views expressed in this presentation are those of the author and do not necessarily represent those of the IMF Execuive Board, Management, or staff.

Page 2: Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

The papers

Hill and Karlsson: Producing a Handbook on Statistics on Rural Development and Agricultural Household Income

Ward: FAO Norms and Standards Gong, Kasnakoglu, and Som: Contributing

to International Classification of Agricultural Products

Wong, Tian, and Sun: On the Research of Statistical Scope of International Agricultural Trade

Page 3: Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

Themes

Hill and Karlsson: How a new Handbook on Statistics on Rural Development and Agricultural Household Income was written under the auspices of the Intersecetariat Working Group on Agricultural Statistics (IWG-Agri) and how it will be updated by the newly formed Wye City Group

Ward: FAO’s strategy to implement statistical standards and build country statistical capacity to report data conforming to them

Gong, Kasnakoglu, and Som: Applying the new international classification of agricultural products in the Central Product Classification (CPC) to the World Program for the Census of Agriculture 2010 (WCA 2010) and the new FAOSTAT data collection questionnaire

Wang, Tian, and Sun: Comparing and contrasting The United States Congress Uruguay Round Agreements Act of 1994 (URAA)

definition of agriculture Agricultural products sections of the Harmonized Commodity Description and

Coding System (HS) The FAO agricultural definition based on the Standard International Trade

Classification (SITC), The UNCTAD agricultural definition, and The International Trade Statistics (ITS) agricultural definition of WTO.

Page 4: Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

Session 7.1 is really about Structural metadata

Structural metadata are collections of codelists Codelists are classifications, such as ISIC, CPC,

SITC, SNA institutional sector, SNA transaction code, SNA asset code, reference period, etc.

Specific instances of the codes from the codelists allow location of a number or text fact within a database “hypercube”

Instances would be an ISIC code for forestry, a CPC or SITC or HS code for cereals and grains, the SNA nonfinancial corporations or households sector, etc.

Page 5: Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

Closely related: Session 7.2 on Referential metadata

Dion: Statistics Canada’s Quality Assurance Framework, Policy on Informing Users of Data Quality and Methodology, and Integrated Meta Data Base

Gong, Som, and Kasnakoglu: FAO’s Annotated Outline for Preparing Country Reports on Metadata for National Agricultural Statistics

Lizarondo and Jalisan: The Philippines CountrySTAT as an extension of FAOSTAT

Related frameworks: IMF’s Data Quality Assessment Framework (DQAF) used in

assessments of statistical system quality called Data Modules of the Reports on Observance of Standards and Codes (data ROSCs)

The Common Metadata Concepts of the Statistical Data and Metadata eXchange (SDMX) standards for computer to computer data communications

Page 6: Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

How metadata relates to comparability

Comparability is critical in the fitness for use of data, particularly when used in international and inter-regional comparisons.

Distinct databases (e.g., on two or more countries) or collections of variables within distinct databases with common structural metadata are comparable

Referential metadata linked to the structural metadata provide additional information on the quality of number or text facts from distinct databases that have the same structural metadata.

Page 7: Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

The importance of metadata standards in data communications

Computer to computer communication between distinct databases (e.g., on two or more countries) requires that the databases be comparable (e.g., have the same structural metadata surrounding the number and text facts they contain)

OR, that their contents be capable of translation into a common “transport” form defined by a standard structural metadata specification

Example: The SDMX data communication protocols now under development by a consortium comprising BIS, ECB, Eurostat, IMF, OECD, UNSD, and World Bank, and implemented by UNSD in disseminating data from its COMTRADE merchandise trade

database The New York Federal Reserve Bank in disseminating exchange rates The FAO in an African pilot study of computer to computer agricultural

data reporting for six West African countries The BIS, OECD, IMF, and World Bank in disseminating, and, ultimately

in exchanging data on external debt statistics

Page 8: Parallel session 3.2: Data Quality & Comparability - System Kim Zieschang IMF Fourth International Conference on Agricultural Statistics 22 October 2007

The importance of metadata standards in data communications

Standard classifications (structural metadata) are important in the fitness for international use of national agricultural data.

The potentially enormous efficiencies of computer to computer data communications in data dissemination and data capture may well provide a last boost to widespread implementation of common classification standards through. e.g., the SDMX data communications protocol.