slide1 n.
Skip this Video
Download Presentation
Jitka Prokop, Czech Statistical Office

Loading in 2 Seconds...

play fullscreen
1 / 17

Jitka Prokop, Czech Statistical Office - PowerPoint PPT Presentation

  • Uploaded on

Jitka Prokop, Czech Statistical Office. SMS-QUALITY T he project and application focused on metadata on quality. European conference on Quality in Official statistics 3-5 June 2014, Vienna. Introduction - aims, features, coverage, current state

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
Download Presentation

PowerPoint Slideshow about 'Jitka Prokop, Czech Statistical Office' - joy

Download Now An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Jitka Prokop, Czech Statistical Office


The project and application

focused onmetadata on quality

European conference on Quality in Official statistics

3-5 June 2014, Vienna

Introduction - aims, features, coverage, current state
  • Architecture - Q-attributes, hierarchical structure, design, preparation, data retrievals and inputs, functionality, stability of values and updates
  • Benchmarking
  • Challenges
  • Topics
Starting points
  • Horizontal way of management
  • Demands for quality reporting & relevant metadata ‚standardisation‘
  • Standardisation of quality reports & adjustments for domain statistics

Tool for managers

  • Semi-interactive cross-cutting overviewsabout quality of a survey (incl. assessments…)
  • Quality reporting
  • Application using web-browserenvironment
  • Reasons and Aims
Q reporting focused on a survey (of any kind) and groups of surveys
  • Cross-cutting info on quality of statistical process, outputdata and products
  • Data preferably retrieved from other source databases
  • Data monitoring, comparisons, aggregations, assessment, benchmarking
  • Flexibilityof metadata content and possibility of survey’s adjustments
  • Help to improve quality reporting and statistical quality itself
  • Encourage self-assessment, support auditing
  • Aims and main features
The application is integrated within the internal SIS and SMS systems
  • Meta-data values retrieved from databases or manually inputted
  • Design & preparation of various quality reports
  • Hierarchical structure (refers to ESQR, GSBPM, DESAP)
  • Publicvs. non-public – individual items or complete reports
  • Bilingual (multi-lingual) solution
  • Usual output formats PDF, HTML, XLS, DBF, DOC, not SDMX
  • User roles: admin, owner, editors, viewers (public vs. internal)
  • Features
Other SMS subsystems can provide certain knowledge on quality criteria e.g. accuracy, relevance, accessibility, clarity, timeliness, punctuality.
  • SMS SURVEYS: statistical processes (particular surveys)
  • SMS REQUIREMENTS: management of main user requirements
  • SMS DISSEMINATION, CATALOGUE OF PUBLICATIONS: dissemination, product quality, info service

in some cases in relation to concrete surveys

  • Interlinks with other SMS-applications
Any statistical process processed or at least with its data stored in the central DWH.
  • Business statistics
  • Social and demography statistics
  • National accounts
  • Price statistics
  • Administrative data statistics.
  • Coverage
Type of info on quality - quantitative and qualitative:
  • Reference metadata
  • Info about process and its phases
  • Schedules
  • Quality performance indicators
  • Calculations
  • Benchmark results
  • Evaluation, (self-)assessments, commentaries
  • Textual, Numerical, Date
  • Q-attribute (item, meta-information, indicator)
Basic information (about a survey)
  • User requirements agenda
  • Methodology info
  • Time schedules; Timeliness; Punctuality
  • Statistical process phases
  • Data confidentiality and protection
  • Data sources; Frame; Sample
  • Outputs and dissemination
  • Individual quality criteria (i.e. quality dimensions)
  • Quality performance indicators
  • Categories of Q-attributes (info on…)
Relates to functionality (stability of values)
  • General
  • Statistical survey (key users, methodology, key statistical variables…)
  • Reference year
  • Processing (all reference periods processed or revised at one time)
  • Reference periods
  • Levels of Q-attributes
To provide relevant & up-to-date information
  • Validity for certain years, batches (i.e. processings), ref.periods
    • When generating data for new reference periods...
  • Metadata updates on each level
    • How to update the derived Q-Maps
    • Managers informed and decide via the application
  • Keeping history and updates
  • Updates of metadata structures and values
Structure (hierarchy): Sections, Sub-sections, Q-attributes
  • Q-Maps: monitoring, benchmarking

General - > Specific -> Survey Q-Maps design, specifications

Value Q-Maps output report

    • Q-Forms: also comparisons and aggregations…

General QM for Q-Forms -> Q-Form -> Value Q-Form

Q-Forms use (not only) ValueQ-Maps as the source of data

  • Q-Maps & Q-Forms
Comparisons, aggregations over
  • Statistical variables
  • Reference periods
  • Surveys
  • Years…

Which data

  • Values
  • Benchmark results
  • Q-Forms - comparisons, aggregations
Design of a report
  • General Q-Map -> A type of report.

General design, pre-setting of parameters.

  • Specific Q-Map -> A group of surveys.

Selection of Q-attributes, way of benchmarking.

  • Survey Q-Map -> A survey

Statistical variables, benchmark scales, links to data.

Output report

  • Value Q-Map -> One reference period.

Retrieval, editing, approval of values. Benchmarking.

  • Levels of Q-Maps - Hierarchy
Primarily for internal management purposes
  • Benchmarked values: numerical or textual
  • Adjustments of scales (boundaries) for particular surveys
  • Parameters
    • To benchmark or not to benchmark?
    • Manually (each value individually) or Automatically (pre-definitions)
    • Categories’ definition –number of categories, and either definition of boundaries or assignment of values from a nomenclature
    • Categories’ labelling –from a special nomenclature or directly in the app
  • Benchmarking
Deeper relations between subsystems
  • Revisions of quality attributes
  • Involvement of domain statisticians
  • Full implementation
  • ESS standard quality reports in SDMX
  • Challenges
Thank you for your attention.

Any questions?