High Level Workshop 2012 Leadership in Modern Statistical Systems Strategies for data collection and data capture. Session 4 - PowerPoint PPT Presentation

Loading...

PPT – High Level Workshop 2012 Leadership in Modern Statistical Systems Strategies for data collection and data capture. Session 4 PowerPoint presentation | free to download - id: 55c93c-ZTIzM



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

High Level Workshop 2012 Leadership in Modern Statistical Systems Strategies for data collection and data capture. Session 4

Description:

High Level Workshop 2012 Leadership in Modern Statistical Systems Strategies for data collection and data capture. Session 4 Olav Ljones Deputy Director General – PowerPoint PPT presentation

Number of Views:108
Avg rating:3.0/5.0
Slides: 17
Provided by: akin80
Learn more at: http://isi.cbs.nl
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: High Level Workshop 2012 Leadership in Modern Statistical Systems Strategies for data collection and data capture. Session 4


1
High Level Workshop 2012Leadership in Modern
Statistical SystemsStrategies for data
collection and data capture. Session 4
  • Olav Ljones
  • Deputy Director General
  • Statistics Norway

2
1.The scope of the session
  • The different data sources, surveys,
    administrative data and registers
  • Legal matters for collection of data
  • Mandatory, even with fines for non response
  • Developments in data capture techniques making
    the most of the new technologies for collection
    and quality control
  • Aligning data collection strategies to the
    available sources in each country.
  • Without data the statistics stop.

3
Reservations
  • Excuse for a perspective built on Norwegian
    experiences.
  • A narrow perspective priority to first step
    data collection. Leave out topics on moving data
    into production systems in NSP, data entry and
    scanning is not discussed.

4
The session - more
  • Official statistics is seen as a production
    process
  • Basic data primary data are the input in the
    production process
  • The theories and techniques of official
    statistics professionally important to
    understand
  • The use of samples instead of full count
    statistics
  • Data collection implies costs for the NSO. Cost
    effiency is important
  • The legal base is more and more important

5
The process The GSPBM
  • http//www.ssb.no/english/subjects/00/90/doc_20081
    7_en/doc_200817_en.pdf

6
One of the important interfaces with the public
  • The response burden
  • Confidentiality a promise
  • Mandatory surveys
  • ..even with fines for non response (Strong power
    given to NSO - DG

7
2. The fundamental strategy for data collection
  • Identify the population of statistical units and
    identify what kind information (set of variables)
    that is requested
  • Examine if reuse of data is possible.
  • If other data sources exists (within the
    government) use administrative data
  • Start the planning of new data collection,

8
The optimal solution- combine
  • resources spent in the NSO
  • the response burden put on respondents
  • statistical quality (accuracy, relevance,
    timeliness etc)
  • Move away from the stove pipe tradition
  • Within the legal and political framwork

9
3. The statistical unit
  • Individuals or
  • Economic units
  • Government
  • Local
  • Central

10
4. Administrative sources
  • Legal rights to use administrative data both
    micro data and accounts and other aggregates.
  • Technical access to data electronic.
    Coordination of electronic communication is
    important.
  • Some influence on planning of the administrative
    data systems e.g. Identification of units,
    concepts and classifications.
  • Rules of confidentiality. The one way stream of
    micro data into statistics. The output from NSO
    is statistics not extended micro files for
    administrative purposes

11
Adm sources cont.
  • Quality checks are important. The European debate
    about statistics and the recent crisis has
    identified lack of quality on the administrative
    data (e.g. public debt) is it acceptable for the
    NSO to act as the public audit authority?
  • Access to administrative data for statistics, but
    without legal control for instance relevant when
    the NSO will give the research community access
    to micro data also from administrative sources.

12
Adm sources more
  • The content of administrative variables has to
    equal the requested statistical variables. This
    strong claim has in practice to be moderated. By
    transformation of the registered variables the
    statistical variables have to be reached. This is
    perhaps the main and most difficult precondition.

13
5. Collection of individual data
  • full count (census) or
  • sample surveys
  • population register
  • increasing non response a problem actions to
    be taken
  • Broaden the common understanding of value of
    official statistics
  • Establish public confidence about the NSO
    respecting fully the confidentiality principles
  • The best modality for the reporting
  • Self reporting e.g. the traditional paper
    questionnaire or is the time ready for web
    email reporting
  • Interviewer - home visit or by phone.
  • The design of questionnaire. Friendly and
    readable/understandable questions
  • Incentives for response and eventually the
    opposite fines for non response
  • Use other sources than individual reporting, E.g.
    Income and wage data from tax authorities.

14
6. Collection of business data.
  • business register
  • Sampling, strata (size), cut off
  • Electronic reporting may be efficient
  • Integrated data collection, new trends. Smart
    solutions if possible
  • Mandatory reporting and in many cases supported
    by fines for non response

15
7. Reshaping and standardisation. New and recent
ideas in official statistics.
  • Standarisation of processes
  • Functional approach
  • Vision Resource efficiency and also a more
    coherent product
  • Implementation problems ?
  • For more information see UNECE the activities
    of HLG- BAS (High Level Group for Strategic
    Developments in Business Architecture in
    Statistics)
  • http//www1.unece.org/stat/platform/display/hlgbas
    /High-LevelGroupforStrategicDevelopmentsinBu
    sinessArchitectureinStatistics28HLG-BAS29js
    essionid33530EA45EA9CAD45E5A35434FFB99BF

16
Thank You
About PowerShow.com