Photis Stavropoulos AGILIS SA

Slides:



Advertisements
Similar presentations
1 European Conference on Quality in Official Statistics Rome, 8-11 July 2008 Improving the quality and the quality assessment of the Labour Force Survey.
Advertisements

EUROPEAN COMMISSION - EUROSTAT ESSnet on Consistency of Concepts and Methods of business-related Statistics 2010 project on statistical units n°
Giovanna Brancato, Marina Signore Istat Work Session on Statistical Metadata (METIS) Metadata and Quality Indicators Reuse for Quality reporting Geneva,
European Conference on Quality in Official Statistics (Q2010) 4-6 May 2010, Helsinki, Finland Brancato G., Carbini R., Murgia M., Simeoni G. Istat, Italian.
REFERENCE METADATA FOR DATA TEMPLATE Ales Capek EUROSTAT.
ESSnet on the use of administrative and accounts data in business statistics Development of Quality Indicators (WP6) John-Mark Frost (ONS, UK), Humberto.
ESTAT E/1-E/ LFS ad hoc module : 1999 Labour Force Survey ad hoc module on Health & Safety at Work 1) Evaluation and recommendations from the Workshop.
24-25 September 2009Meeting of the European Directors of Social Statistics Update on Lifelong Learning and Education Systems statistics Agenda point 4.8.
13 November, 2014 Seminar on Quality Reports QUALITY REPORTS EXPERIENCE OF STATISTICS LITHUANIA Nadiežda Alejeva Head, Price Statistics.
Session topic (i) – Editing Administrative and Census data Discussants Orietta Luzi and Heather Wagstaff UNECE Worksession on Statistical Data Editing.
Implementation of Quality indicators for administrative data
Weighting issues in EU-LFS
The usage of web interviewing in Lithuanian Labour Force Survey
5. Areas under organic farming
LFS ad hoc module 2009 “Entry of young people into the labour market”
Bettina Wistrom OECD Statistics Directorate
AES progress report and future plans
Survey phases, survey errors and quality control system
Item 5.3 Feasibility studies
Adult Education Survey
Strategic development group Update on education and lifelong learning statistics 17/6/2009.
Education and Training Statistics Working Group 24-25/9/2007
Survey phases, survey errors and quality control system
Item 10 – Conclusions of the meeting
2003 LFS ad hoc module on lifelong learning
Item 4.2 – Implementation of the Adult Education Survey
LAMAS Working Group January 2016
Education and Training Statistics Working Group
Emilio Di Meglio and Emanuela Di Falco (EUROSTAT)
National needs for AES Purpose - describe participation in learning during a 12 months period. The main parameters are; Participation rates in different.
LAMAS October 2017 Agenda Item 3.2 Labour Cost Indices state of play Daniel Iscru Hubertus Vreeswijk.
Passenger Mobility Statistics 10 April 2014
Adult Education Survey : recommendations of the TF AES
Overview of Approaches to Register-Based Populating Censuses
Item 6.2 Participation in education and training in the last 12 months
LAMAS Working Group 29 June-1 July 2016
Item 7.1 Implementation of the 2016 Adult Education Survey
Item 8.1 Implementation of the 2016 Adult Education Survey
State of play: data transmission, validation and dissemination
Education and Training Statistics Working Group – 1-2 June 2017
Education and Training Statistics Working Group Meeting of 17 November 2010 Item 3 – Lifelong learning and outcomes of education Sylvain Jouhette Eurostat.
National Educational Attainment Classifications (NEACs)
Education and Training Statistics Working Group, May 2011
Education and Training Statistics Work programme 2004
ESQRS implementation in the in Labour Force Survey
LAMAS Working Group 6-7 December 2017
Mapping Data Production Processes to the GSBPM
Meeting of the Management Group on Statistical Co-operation
Item 4.3 – Repeal of CVTS legal acts
Education and Training Statistics Working Group Meeting 5/6 June 2012 Item 4.2 Progress report on education survey data and metadata Sylvain Jouhette.
Eurostat ETS Working Group - Luxembourg, April 2013
Item 7.2 – Overview of 2012 UOE Quality Reports
“Education and the labour market” in NewCronos
Education and Training Statistics Working Group – June 2014
Item 4 Overview of the 2016 AES & 2015 CVTS data collection
Item 4.2 – Towards the 2016 AES Philippe Lombardo Eurostat-F5
Adult Education Survey mandate of the TF AES
Education and Training Statistics Working Group, May 2011
Grants for the implementation of ISCO 08 during 2010
COMITOLOGY ITEMS (Point 4. on the agenda)
Item 8 - Disability statistics
LAMAS Working Group October 2018
Quality project regional GVA and employment
Basic preconditions The next round of population and housing censuses is scheduled for the start of the new decade (2021), both in the EU and in the partner.
2.7 Annex 3 – Quality reports
Health / disability variables in the LFS Item 2.10 of the agenda
Directors of Social Statistics 2009 MODULE ON MATERIAL DEPRIVATION
SILC draft implementing and delegated acts Item 3.4 of the agenda
Item 5 Modernisation of the EU-SILC Production
ESS conceptual standards for quality reporting
Presentation transcript:

Photis Stavropoulos AGILIS SA Draft quality report of the implementation of the 2003 LFS ad hoc module on lifelong learning Photis Stavropoulos AGILIS SA ETS-WG meeting, 31/01-02/02, 2005

Introduction 30 countries included an ad hoc module on Lifelong Learning (LLL) in the 2003 Labour Force Survey (LFS) A draft quality report about the implementation of the module has been prepared, based on information provided by the countries until 7/1/2005 This talk presents the main findings contained in the draft report The final report is scheduled for 31/3

Overview Collection of information about the implementation of the module Main aspects of the implementation of the module Draft quality assessment Planned improvements for the final quality report

Information collection Countries submitted MEGs, Grant reports, Self evaluation reports, etc. A metadata questionnaire was prepared Questionnaires were pre-filled with submitted information and sent back to countries Questionnaires returned until 7/1/2005 and available reports were used for the preparation of the draft quality report

Implementation of the module Adoption of the questionnaire Most variables were adopted as defined in Commission Regulation 1313/2002 A few of them however were either modified or not included at all, e.g. HATFIELD, LLLFIELD were considered covered by HATFIELD and EDUCFIELD of the core questionnaire (ES, FI, SE, DK, HU, IT) Registers were used for LLLLEVEL and LLLFIELD (NO) and for HATFIELD (SE and FI) UK constrained questions about non-regular education to the most recent activity and Norway to the most recent job-related activity Most extended modifications in Norway, UK and France

Implementation of the module Target population Regulation 1313/2002: persons aged 15 years and over Exceptions 15-74: DK, EE, FI, HU, LV, NL SE, 16-74: CH, IS, 16 - … : ES, UK, 15-64: BE Norway used a target population of 15 to 59 years for questions on regular education (registers for older persons) and 16 to 74 for learning outside the regular education system. Sampling scheme Two-stage stratified sampling of individuals or households in the majority of cases

Implementation of the module Mode of data collection PAPI or CAPI, often with CATI Only CATI in the Nordic countries Postal survey (Denmark, in addition to CATI) Proxy interviews allowed in most countries Response rates

Implementation of the module Reported users and uses of the module’s results Users: Ministries, media, academia, research organizations Uses: Indicators on LLL, linkage of information on training activities with information on their cost, comparisons at international level Dissemination of results Publications in paper and electronic format, mailing lists, provision of anonymized microdata to researchers

Implementation of the module Main reported problems Recalling of short term learning activities Inefficient coding of field of training High percentage of “General programs” in field-of education questions Difficulty in distinction between formal and non-formal education High non-response about taught activities Inaccurate information Lack of coherence of results with other sources High cost of the survey

Implementation of the module Suggestions for improvement, made by countries Reduce reference period Limit detailed investigation of taught activities to the most recent one Ask about duration of taught activities in closed form (pre-defined time intervals) Include additional checks during data entry Improve training of interviewers on coding of field of education Provide more explanations in the questionnaire about the questions

Quality assessment Relevance LFS regulation restricts the size of the module Consequence: not all potential needs can be covered Coverage of needs for indicators of LLL Participation rates and time spent in LLL (covered) Perceived interest in learning (partly covered - LLLCOURPURP) Sources of financing and employer support for learning (partly covered - LLLCOURWORH) Outcome of learning activities (not covered) Barriers to learning (not covered) Investment in LLL (not covered) Conformity of countries with Commission Regulation Modification or non-inclusion of certain variables Reduction of target population in Belgium (15-64)

Quality assessment Accuracy Direct assessment (variability of results) Sweden reports rates with s.e. < 1.5% in most cases Latvia reports main rates of participation in LLL with c.v < 5.85% Indirect assessment Coverage errors: reported at below 2%, except EE (5%) and ES (17%) Measurement error sources: difficulty to distinguish formal from non-formal education, comprehension problems for informal training activities, long reference period, proxy interviews, coding errors during the interview Processing errors: Coding errors at the statistical agency Non-response: the rate of non-response was non-negligible in several countries, no information about differences between respondents and non-respondents

Quality assessment Comparability Concepts Measurement processes Variables used: Countries with greatest deviation from the regulation were France, UK and Norway Statistical unit: Individual and educational activity (common for all countries) Target population: Major discrepancies were found in Belgium(15-64) Reference period of results: Sweden, France and Iceland implement the module throughout 2003, extending the reference period of results Measurement processes Various sampling schemes, not common data collection mode, difficulties in education field coding

Quality assessment Coherence HATFIELD is in agreement with data from administrative registers in Denmark LLLLEVEL compared with data from educational institutes in Bulgaria Overestimates participants up to and including upper secondary education  due to different reference periods Underestimates participants to higher education  difference possibly within limits of sampling error LLLCOURWORH compared with Staff Training Survey (STS) in Sweden Underestimates training financed by the employers Consistent patterns of participation in employer-financed activities by sex and level of education Inconsistent differences in rates between subpopulations LLLLEVEL, LLLCOURLENP, LLLCOURPURP incoherent with core LFS questionnaire

Quality assessment Timeliness and punctuality Distance between end of reference period and time of delivery to Eurostat ranges from 6 to 11 months Punctuality: Sweden delivered data 2 months after scheduled time, Hungary was on time

Accessibility and clarity Quality assessment Accessibility and clarity Users can obtain statistics in the form of results publications (paper and electronic) and anonymized microdata (Ireland, Hungary, Latvia) Methodological information (Ireland and Hungary) Information in MEGs, Grant reports, self evaluation reports  not widely available

Planned improvements for final quality report Presentation of data processing and analysis Detailed description on data editing, treatment of non-response, presentation of the results Problems and suggestions The section will be revised with new information Relevance More information on user needs, users and uses of the results of the module Accuracy Coefficients of variation and non-sampling errors Comparability Visual representations of comparability using multidimensional scaling and clustering techniques on the complete metadata Coherence Comparison of the EU results and similar EU-wide results

… your comments are welcome