metadata: integral part of statistics canada quality framework international conference on...

20
Metadata: Integral Part of Statistics Canada Quality Framework International Conference on Agriculture Statistics October 22-24, 2007 Marcelle Dion Director General, Agriculture, Technology and Transportation Statistics, Statistics Canada

Upload: luke-harrell

Post on 26-Dec-2015

212 views

Category:

Documents


0 download

TRANSCRIPT

Metadata: Integral Part of Statistics Canada Quality Framework

International Conference on Agriculture Statistics

October 22-24, 2007

Marcelle DionDirector General, Agriculture, Technology and Transportation Statistics, Statistics Canada

Outline

Quality Assurance Management

Integrated Metadata Base

Application to Agriculture Statistics

Statistics Canada’s Quality Assurance Management

Three key elements:Quality Assurance Framework

Policy on Informing Users of Data Quality and Methodology

Integrated Metadata Base

Quality Assurance Framework

Defines quality as “fitness for use”Identify six indicators of “fitness for use”

Relevance, accuracy, timeliness, accessibility, interpretability and coherence

Metadata: at the heart of the management of the interpretability indicator

Interpretability refers to the provision of information that help users understand the data released by Statistics Canada

Policy on Informing Users of Data Quality and Methodology

One of the policies that governs Statistics CanadaRequires all statistical outputs to have information (metadata) on:

Concepts and definitionsMethodology used to produce the dataData accuracy

Defines standards and guidelinesIdentifies responsibilities

Integrated Metadata Base

Definition of Metadata

Characteristics of the Data Base

GovernancePolicy

Technical Assistance

Monitoring

Access

Definition of Metadata

Metadata inform users of the features that affect the quality of all data published by Statistics Canada. provide a better understanding of the strengths and limitations of data, and how they can be effectively used and analyzed are of particular importance when making comparisons with

data across surveys or sources of informationin drawing conclusions regarding

changes over timedifferences between geographic areas differences among sub-groups of the target populations of surveys

breed users’ trust

Integrated Metadata BaseCharacteristics

Corporate repository of metadata

Metadata for 415 active and 400 inactive surveys (discontinued or one time)

Metadata for any survey instance since November 2000

Easily accessible online: www.statcan.ca

Metadata complies with policyMinimum requirements

Metadata: Policy’s Minimum Requirements

Integrated Metadata BaseGovernance

Policy ensures compliance

Policy ensures coherence

Policy defines accountabilityManagers of program areas

Methods and Standards Committee

Standards Division responsible for management

Integrated Metadata BaseTechnical Assistance

Metadata Officers

Guidelines for author

Template

Template – One SectionData Sources – Type of Surveys

Please highlight the terms that apply to this survey (more than one source may apply)Direct – data are collected directly from STC respondents with the use of a collection instrumentAdministrative - data are extracted from administrative files provided by an external organization that collected them for reasons other than statistical purposesDerived – Data were derived from other Statistics Canada programs or surveys and/or other sources (e.g. media, annual reports)

---------------------------------------------------------Census – the intent is to collect information from all units of the survey populationSample – information is collected from only a fraction of units of the survey population ----------------------------------------------------------Longitudinal – the same statistical units are followed over timeCross-sectional – the statistical units are specific to one point in timeCross-sectional with longitudinal follow-up – the statistical units which are specific to one point in time are also followed over time -----------------------------------------------------------Mandatory – respondents are required, under the Statistics Act, to answer all the survey questionsVoluntary – respondents may refuse to answer some or all of the survey questions

Integrated Metadata BaseQuality Monitoring

Triggers for creating or updating records

Information loading: Template

Metadata officers’ review Identify program areas needing assistance

Exhaustive review - Four-point scale

Official notice to program areas’ managers

Users’ AccessMetadata Flow

STATISTICAL PROGRAMS

IMDB Team in Standards Division

Integrated Metadatabase•Definition of variable

•Survey description/methodology/questionnaire/documentation•Data accuracy

Web Page Generation of Documentation

¤ CANSIM ¤ Summary Table ¤ Analytical Studies¤ Online Catalogue ¤ Publication ¤ Information for Survey Participants¤ The Daily ¤ Definitions, Data Sources and Methods

Agriculture Statistics

Description of the program

Framework

Metadata

Program Description

Monthly, quarterly, annual and/or seasonal data collection activities related to crop and livestock and farm finances as needed Quinquennial Census of Agriculture with the Census of PopulationEconomic series on the agriculture sector

System of National Accounts (SNA) and agriculture GDPAdministrative data supplement limited survey taking in supply-managed agriculture sectors such as dairy and poultryTaxation data for annual disaggregated financial informationCost-recovery program (e.g. farm assets and liabilities, etc.)Farm Register

Agriculture surveys’ frame Large, complex agricultural operations’ profile

Joint collection agreements with most provincial and territorial departments of agriculture

Agriculture Statistics Framework

Agriculture Program Metadata

IMDB includes 45 separate records covering current survey activities One step further: the statistical activity

IMDB structure that groups together surveys that share common processing system or conceptual framework

Agriculture statistics program statistical activitywould be organized around the farm income series structure would facilitate users’ understanding of the interrelationships between the various components of the integrated program Would increase users’ awareness that the “fitness for use” test of the farm income series should take into account the metadata information of all the farm income series’ inputs.

Conclusion

Data quality is a survival issue for any statistical agency

loss of users’ confidence in data would render statistical agency ineffective

Metadata: openness and transparency about data –about their weaknesses just as much as about their strengthsIMDB – by making metadata easily accessible to users – helps build trust

For more information Pour plus de renseignementsplease contact veuillez contacter

Marcelle DionDirector General Agriculture, Technology

and Transportation Statistics BranchStatistics Canada

13th Floor Section B-7, Jean Talon Building170 Tunney’s Pasture Driveway

Ottawa (Ontario), CanadaK1A 0T6

Email: [email protected]

www.statcan.ca