Survey phases, survey errors and quality control system

Slides:



Advertisements
Similar presentations
ESRC UK Longitudinal Studies Centre A Framework for Quality Profiles Nick Buck and Peter Lynn Institute for Social and Economic Research University of.
Advertisements

Preparing Data for Quantitative Analysis
An integrated information system on survey quality: the experience of the Italian survey ‘Holidays and trips’ by Monica Perez 7th International Forum on.
Metadata to Support the Survey Life Cycle Alice Born, Statistics Canada Joint UNECE/Eurostat/OECD Work Session on Statistical Metadata (METIS) Geneva,
Quality Guidelines for statistical processes using administrative data European Conference on Quality in Official Statistics Q2014 Giovanna Brancato, Francesco.
Documentation and survey quality. Introduction.
United Nations Economic Commission for Europe Statistical Division Applying the GSBPM to Business Register Management Steven Vale UNECE
Environment Change Information Request Change Definition has subtype of Business Case based upon ConceptPopulation Gives context for Statistical Program.
Arun Srivastava. Types of Non-sampling Errors Specification errors, Coverage errors, Measurement or response errors, Non-response errors and Processing.
Giovanna Brancato, Marina Signore Istat Work Session on Statistical Metadata (METIS) Metadata and Quality Indicators Reuse for Quality reporting Geneva,
Marina Signore Head of Service “Audit for Quality Istat Assessing Quality through Auditing and Self-Assessment Signore M., Carbini R., D’Orazio M., Brancato.
The Adoption of METIS GSBPM in Statistics Denmark.
Eurostat Overall design. Presented by Eva Elvers Statistics Sweden.
Current and Future Applications of the Generic Statistical Business Process Model at Statistics Canada Laurie Reedman and Claude Julien May 5, 2010.
European Conference on Quality in Official Statistics Session 26: Quality Issues in Census « Rome, 10 July 2008 « Quality Assurance and Control Programme.
United Nations Economic Commission for Europe Statistical Division Mapping Data Production Processes to the GSBPM Steven Vale UNECE
Use of Administrative Data Seminar on Developing a Programme on Integrated Statistics in support of the Implementation of the SNA for CARICOM countries.
MGT-491 QUANTITATIVE ANALYSIS AND RESEARCH FOR MANAGEMENT OSMAN BIN SAIF Session 5.
Supporting Researchers and Institutions in Exploiting Administrative Databases for Statistical Purposes: Istat’s Strategy G. D’Angiolini, P. De Salvo,
Statistical Expertise for Sound Decision Making Quality Assurance for Census Data Processing Jean-Michel Durr 28/1/20111Fourth meeting of the TCG - Lubjana.
Pilot Census in Poland Some Quality Aspects Geneva, 7-9 July 2010 Janusz Dygaszewicz Central Statistical Office POLAND.
Developing and applying business process models in practice Statistics Norway Jenny Linnerud and Anne Gro Hustoft.
United Nations Oslo City Group on Energy Statistics OG7, Helsinki, Finland October 2012 ESCM Chapter 8: Data Quality and Meta Data 1.
Census Processing Baku Training Module.  Discuss:  Processing Strategies  Processing operations  Quality Assurance for processing  Technology Issues.
5.8 Finalise data files 5.6 Calculate weights Price index for legal services Quality Management / Metadata Management Specify Needs Design Build CollectProcessAnalyse.
METIS 2011 Workshop Session III – National Implementation of the GSBPM Alice Born and Tim Dunstan Thursday October 6, 2011 Implementation of the GSBPM.
First meeting of the Technical Cooperation Group for the Population and Housing Censuses in South East Europe Vienna, March 2010 POST-ENUMERATION.
A Training Course for the Analysis and Reporting of Data from Education Management Information Systems (EMIS)
Metadata requirements for archiving structured data Alice Born Statistics Canada Joint UNECE/Eurostat/OECD Work Session on Statistical Metadata (9-11 April.
Session topic (i) – Editing Administrative and Census data Discussants Orietta Luzi and Heather Wagstaff UNECE Worksession on Statistical Data Editing.
Metadata models to support the statistical cycle: IMDB
Chapter 1 Introduction and Data Collection
PROCESSING DATA.
Implementation of Quality indicators for administrative data
Theme (v): Managing change
Generic Statistical Data Editing Models (GSDEMs)
Information for marketing management
Quality assurance in official statistics
Systems Analysis and Design
Market Research Unit 3 P3.
Data Collection Methods
LIVESTOCK PRODUCTION AND PRODUCTIVITY
Data Collection techniques Marina Signore
An Active Collection using Intermediate Estimates to Manage Follow-Up of Non-Response and Measurement Errors Jeannine Claveau, Serge Godbout and Claude.
Sample surveys versus business register evaluations:
Generic Statistical Business Process Model (GSBPM)
Warm up – Unit 4 Test – Financial Analysis
Survey phases, survey errors and quality control system
Tomaž Špeh, Rudi Seljak Statistical Office of the Republic of Slovenia
Organization of efficient Economic Surveys
6.1 Quality improvement Regional Course on
Aurora De Santis, Riccardo Carbini Istat, Italy
Introduction to Quality Concepts
Quality assessment ESTP Training Course “Quality Management and survey Quality Measurement” Rome, 24 – 27 September 2013 Giorgia Simeoni Researcher Unit.
UNODC-UNECE Manual on Victimization Surveys: Content
Jeroen Pannekoek, Sander Scholtus and Mark van der Loo
Measurement errors Marina Signore
Business Statistics: A First Course (3rd Edition)
ENCODING TOOL DEVELOPED BY HUNGARY Márta Záhonyi
Mapping Data Production Processes to the GSBPM
Parallel Session: BR maintenance Quality in maintenance of a BR:
Karin Blix, Statistics Denmark,
GSBPM AND ISO AS QUALITY MANAGEMENT SYSTEM TOOLS: AZERBAIJAN EXPERIENCE Yusif Yusifov, Deputy Chairman of the State Statistical Committee of the Republic.
Indicator 3.05 Interpret marketing information to test hypotheses and/or to resolve issues.
Training course on developing and using questionnaires for agricultural surveys Field Testing Post evaluation methods Marco Ballin Istanbul, July.
2.7 Annex 3 – Quality reports
Multi-Mode Data Collection
Deciding the mixed-mode design WP1
Joint UNECE/Eurostat/OECD
GSBPM Giorgia Simeoni, Istat,
Presentation transcript:

Survey phases, survey errors and quality control system Measurement of the quality of statistics 3-5 October 2012   Marina Signore Istat Division "Metadata, Quality and R&D Projects", Chief

Topics Survey phases and survey errors Quality control system

Survey phases Set of homogenous operations from an organisational and chronological point of view Each phase is a potential source of errors Error sources and methods to prevent or reduce errors impact for each phase of the statistical production process

GSBPM: Generic Statistical Business Process Model

GSBPM main features Developed within the Joint UNECE/Eurostat/OECD Work Session on Statistical Metadata (METIS) Born as a documentation model, with the aim of standardise terminology Organised in four levels with increasing level of detail Over-arching processes Flexibility Final version April 2009

Over-arching processes The Generic Statistical Business Process Model Level 0 Over-arching processes Level 1 Level 2 5.2. Classify and code - This sub-process classifies and codes the input data. For example automatic (or clerical) coding routines may assign numeric codes to text responses according to a pre-determined classification scheme. Level 3

Questionnaire design and testing planning Sample selection Data processing Data collection Documentation Quality control system

Quality control system Survey phases Planning Questionnaire design and testing Sample selection Data collection Data processing Documentation Quality control system

Planning Determine needs for information Design outputs Design variables descriptions Design data collection methodology Design frame and sample methodology Design production system & workflow Design the quality control system Take into account constraints in terms of available resources, costs, timeliness, national or international regulations.

Planning Determine needs for information Design outputs Impact on relevance, comparability and coherence Planning Determine needs for information Design outputs Design variables descriptions Design data collection methodology Design frame and sample methodology Design production system & workflow Impact on accuracy but also on timeliness

Survey phases Questionnaire design and testing Sample selection Data collection Data processing Impact on accuracy but also on timeliness

Questionnaire design and testing The questionnaire is a communication tool highly influential for data quality question wording response categories and their order the way and context in which questions are presented Source of measurement errors and item non response In some cases also of unit non response

Sample selection The selection from a frame of the population units given the sample design and size the population to be investigated should have been specified and the frame from which the units are sampled and contacted should have been chosen Source of coverage or frame errors

Data collection data collection is any process whose purpose is to acquire or assist in the acquisition of data* the impact of data collection is both direct and critical as these data are the primary input. The quality of the operation has a very high impact on the quality of the final product, in particular on accuracy the mode of administration of the questionnaire (e.g. mail, telephone, in person) has an influence on data quality. Respondents may answer questions differently in the presence of an interviewer, over the phone, or by themselves Source of unit non response and measurement errors (respondent and interviewer errors) But also data collection mode effect *Statistics Canada (2003)

Data processing - data entry - editing and imputation - coding Coding: operation of converting an actual response for a survey question into a category (open-ended responses e.g. “other specify” or “specify occupation” Data entry: process of transferring collected data to an electronic medium Editing and Imputation: data editing is the application of checks to detect missing, invalid or inconsistent entries or to point to data records that are potentially in error and the imputation is the process used to determine and assign replacement values for missing, invalid, inconsistent data that have failed edits

Data processing Source of processing errors the errors range from simple recording errors to mispecification of the editing and imputation model. different impact for different data collection mode. e.g. different coders are likely to interpret and code the same response differently. Even the same coder may change over time (get experienced or bored). They affect the accuracy of final estimates Source of processing errors

Survey phases and survey errors Questionnaire design Measurement errors Item non response Sample selection Coverage or frame errors Data collection Unit non response Data processing Processing errors

Quality Control System Set of tools and methods to improve data quality Set of quality control actions for improving the quality of each survey phase and operations Quality control activity should be part of the survey process itself

Quality Control Actions Preventive actions: actions put in place in order to avoid potential errors usually before the operation is carried out Monitoring actions: actions put in place in order to reduce errors during the execution of an operation Ex-post actions: actions put in place in order to estimate the impact on the final data of non sampling errors

Some examples Quality control system Training of interviewers Advance letter to respondents Interviewers response rate Questionnaire testing Follow-up on non respondent units Edit and imputation procedures Sampling variance estimation

Quality Control System Planning quality Prevent Monitor Evaluate Run the checks Process Indicators Direct Measures Analyse results

Quality Measures Direct measures Quality indicators + estimate of error impact - very expensive - no overall estimate Quality indicators - indirect estimate of error impact + by-product of the production process + alarm bell

Documentation It is important to document the survey operations and procedures, the quality controls done and their results It is a demanding and time-consuming task and usually has not a high priority for survey managers

Documentation It should be considered as an investment: Save time in future survey occasions Easier to train new staff Share good practices and know-how Transparency

References Biemer P. and Lyberg L. (2003), Introduction to Survey Quality, J. Wiley, N.Y. Groves R.M. (1989), Survey Errors and Survey Costs, J. Wiley, N.Y. Lyberg L. and Kasprzyk D. (1991), “Data Collection Methods and Measurement Errors: An Overview” in Measurement errors in surveys, Biemer et alt. (eds.), J. Wiley, N.Y. Lyberg L. and Kasprzyk D. (1997), “Some Aspects of Post-Survey Processing” in Survey Measurement and Process Quality, Lyberg et alt. (eds.), J. Wiley, N.Y. ONS (2004), Guidelines for Measuring Statistical Quality, version 1.0, http://www.statistics.gov.uk/methods_quality/publications.asp Statistics Canada (2003), Quality Guidelines, fourth edition 25 http://www.statistics.gov.uk/methods_quality/publications.asp 25