Stop the Madness: Use Quality Targets Laurie Reedman.

Slides:



Advertisements
Similar presentations
Data Quality Assurance and Dissemination International Workshop on Energy Statistics Aguascalientes, Mexico.
Advertisements

Evaluating the Effects of Business Register Updates on Monthly Survey Estimates Daniel Lewis.
Quality Review of Key Indicators at Statistics Canada ICES-III, June 2007 Claude Julien and Don Royce.
Innovation Surveys: Advice from the Oslo Manual South Asian Regional Workshop on Science, Technology and Innovation Statistics Kathmandu,
Innovation Surveys: Advice from the Oslo Manual National training workshop Amman, Jordan October 2010.
Group Work on Indicators Objective: To develop a list of priority indicators (both results and process) for each intervention point for further refinement.
Wisconsin Knowledge & Concepts Examination (WKCE) Test Security Training for Proctors Wisconsin Department of Public Instruction Office of Educational.
The estimation strategy of the National Household Survey (NHS) François Verret, Mike Bankier, Wesley Benjamin & Lisa Hayden Statistics Canada Presentation.
Enhancing Data Quality of Distributive Trade Statistics Workshop for African countries on the Implementation of International Recommendations for Distributive.
Workshop on Energy Statistics, China September 2012 Data Quality Assurance and Data Dissemination 1.
Kevin Deardorff Assistant Division Chief, Decennial Management Division U.S. Census Bureau 2014 SDC / CIC Conference April 2, Census Updates.
1 Editing Administrative Data and Combined Data Sources Introduction.
Aaker, Kumar, Day Ninth Edition Instructor’s Presentation Slides
OECD Short-Term Economic Statistics Working PartyJune Analysis of revisions for short-term economic statistics Richard McKenzie OECD OECD Short.
08/08/2015 Statistics Canada Statistique Canada Paradata Collection Research for Social Surveys at Statistics Canada François Laflamme International Total.
Energy Efficiency Benchmarking for Mobile Networks
1 Module 6 Putting It All Together. 2 Learning Objectives At the end of this session participants will understand: The monitoring and evaluation process.
André Loranger New York, June 2014 The Integrated Business Statistics Program at Statistics Canada Presentation to the UNCEEA Assistant Chief Statistician.
Community Planning Training 1-1. Community Plan Implementation Training 1- Community Planning Training 1-3.
Edit and Imputation of the 2011 Abu Dhabi Census Glenn Hui and Hanan AlDarmaki Statistics Centre - Abu Dhabi UNECE CES Work Session on Statistical Data.
The Canadian Census of Population: a Review in Preparation for 2016 UNECE Group of Experts on Population and Housing Censuses May 23, 2012.
18/08/2015 Statistics Canada Statistique Canada Responsive Collection Design (RCD) for CATI Surveys and Total Survey Error (TSE) François Laflamme International.
Vienna, 23 April 2008 UNECE Work Session on SDE Topic (v) Editing on results (post-editing) 1 Topic (v): Editing based on results Discussants: Maria M.
1 Early Childhood Special Education Connecticut State Department of Education Early Childhood Special Education Maria Synodi.
Aggregate and Systemic Components of Risk in Total Survey Error Models John L. Eltinge U.S. Bureau of Labor Statistics International Total Survey Error.
RESEARCH A systematic quest for undiscovered truth A way of thinking
Giovanna Brancato, Marina Signore Istat Work Session on Statistical Metadata (METIS) Metadata and Quality Indicators Reuse for Quality reporting Geneva,
Volunteer Angler Data Collection and Methods of Inference Kristen Olson University of Nebraska-Lincoln February 2,
Use of survey (LFS) to evaluate the quality of census final data Expert Group Meeting on Censuses Using Registers Geneva, May 2012 Jari Nieminen.
Q2010, Helsinki Development and implementation of quality and performance indicators for frame creation and imputation Kornélia Mag László Kajdi Q2010,
Use of Administrative Data in Statistics Canada’s Annual Survey of Manufactures Steve Matthews and Wesley Yung May 16, 2004 The United Nations Statistical.
Assessing Quality for Integration Based Data M. Denk, W. Grossmann Institute for Scientific Computing.
Eurostat Overall design. Presented by Eva Elvers Statistics Sweden.
Deliverable 2.6: Selective Editing Hannah Finselbach 1 and Orietta Luzi 2 1 ONS, UK 2 ISTAT, Italy.
CZECH STATISTICAL OFFICE 1 The Quality Metadata System In the Czech Statistical Office Work Session on Statistical Metadata (METIS)
A Strategy for Prioritising Non-response Follow-up to Reduce Costs Without Reducing Output Quality Gareth James Methodology Directorate UK Office for National.
The application of selective editing to the ONS Monthly Business Survey Emma Hooper Office for National Statistics
Jeroen Pannekoek - Statistics Netherlands Work Session on Statistical Data Editing Oslo, Norway, 24 September 2012 Topic (I) Selective and macro editing.
European Conference on Quality in Official Statistics Session 26: Quality Issues in Census « Rome, 10 July 2008 « Quality Assurance and Control Programme.
PPA 502 – Program Evaluation Lecture 2c – Process Evaluation.
Statistik.atSeite 1 Norbert Rainer Quality Reporting and Quality Indicators for Statistical Business Registers European Conference on Quality in Official.
A Theoretical Framework for Adaptive Collection Designs Jean-François Beaumont, Statistics Canada David Haziza, Université de Montréal International Total.
Quality Assurance Programme of the Canadian Census of Population Expert Group Meeting on Population and Housing Censuses Geneva July 7-9, 2010.
Statistics Canada Statistique Canada Cost-Efficient Framework for Data Collection for CATI Surveys Social Surveys Collection Research Steering Committee.
European Conference on Quality in Official Statistics 8-11 July 2008 Mr. Hing-Wang Fung Census and Statistics Department Hong Kong, China (
Statistical Expertise for Sound Decision Making Quality Assurance for Census Data Processing Jean-Michel Durr 28/1/20111Fourth meeting of the TCG - Lubjana.
A Quality Driven Approach to Managing Collection and Analysis
Topic (iii): Macro Editing Methods Paula Mason and Maria Garcia (USA) UNECE Work Session on Statistical Data Editing Ljubljana, Slovenia, 9-11 May 2011.
Outlining a Process Model for Editing With Quality Indicators Pauli Ollila (part 1) Outi Ahti-Miettinen (part 2) Statistics Finland.
Integrated Approach Processing Marie Brodeur Director General, Industry Statistics Branch, Statistics Canada St. Lucia February, 2014 SNA seminar in the.
Topic (i): Selective editing / macro editing Discussants Orietta Luzi - Italian National Statistical Institute Rudi Seljak - Statistical Office of Slovenia.
United Nations Oslo City Group on Energy Statistics OG7, Helsinki, Finland October 2012 ESCM Chapter 8: Data Quality and Meta Data 1.
Multivariate selective editing via mixture models: first applications to Italian structural business surveys Orietta Luzi, Guarnera U., Silvestri F., Buglielli.
QUALITY ASSESSMENT OF THE REGISTER-BASED SLOVENIAN CENSUS 2011 Rudi Seljak, Apolonija Flander Oblak Statistical Office of the Republic of Slovenia.
Proposal on Revised Mechanism of Selecting Applications for Approval Presentation by Secretariat of Council for the AIDS Trust Fund in Sharing Session.
Copyright 2010, The World Bank Group. All Rights Reserved. QUALITY ASSURANCE AND EVALUATION Part 1: Quality Assurance 1.
Testing the use of administrative data to edit the 2009 Agriculture Census Dolores Lorca National Statistical Institute of Spain.
The development of a data editing and imputation tool set UN/ECE Work Session on Statistical Data Editing Topic (ii): Global solutions to editing Claude.
Monitoring and evaluation Objectives of the Session  To Define Monitoring, impact assessment and Evaluation. (commonly know as M&E)  To know why Monitoring.
Demonstrating Institutional Effectiveness Documenting Using SPOL.
Monitoring and Evaluation Systems for NARS Organisations in Papua New Guinea Day 3. Session 9. Periodic data collection methods.
Étienne Saint-Pierre and Serge Godbout, Statistics Canada
An Active Collection using Intermediate Estimates to Manage Follow-Up of Non-Response and Measurement Errors Jeannine Claveau, Serge Godbout and Claude.
Survey phases, survey errors and quality control system
Measuring Data Quality and Compilation of Metadata
Survey phases, survey errors and quality control system
“Managing Modern National Statistical Systems in Democratic Societies”
Assessing Quality of Paradata to Better Understand the Data Collection Process for CAPI Social Surveys François Laflamme Milana Karaganis European Conference.
A modest attempt at measuring and communicating about quality
Étienne Saint-Pierre, Statistics Canada
Presentation transcript:

Stop the Madness: Use Quality Targets Laurie Reedman

Scope  Aspects of quality Timeliness and accuracy  Mechanisms to manage quality Indicators and pre-set targets  Survey processes Computer assisted interviewing Collection follow-up Manual processing 25/10/2015 Statistics Canada Statistique Canada 2

Statistics Canada’s dimensions of quality  Relevance  Accuracy  Timeliness  Accessibility  Interpretability  Coherence 25/10/2015 Statistics Canada Statistique Canada 3

Statistics Canada’s dimensions of quality  Relevance  Accuracy  Timeliness  Accessibility  Interpretability  Coherence 25/10/2015 Statistics Canada Statistique Canada 4

Statistics Canada’s dimensions of quality  Relevance  Accuracy  Timeliness  Accessibility  Interpretability  Coherence 25/10/2015 Statistics Canada Statistique Canada 5 How can a survey manager manage both process and product quality of data collection and manual processing activities?

Interviewer Monitoring  Computer assisted interviewing  Monitor observes and grades samples of interviewer work  Frequency of monitoring sessions geared to attain desired average outgoing quality level 25/10/2015 Statistics Canada Statistique Canada 6

Responsive Collection Design (RCD)  An adaptive approach to survey data collection  Uses information prior to and during data collection to adjust the strategy for the remaining in-progress cases (Groves and Herringa, JRSS 2006)  Can use RCD to: Control quality (response rate, representativeness) Control cost (time and resources spent) 25/10/2015 Statistics Canada Statistique Canada 7

Responsive Collection Design (RCD)  RCD was piloted on 2 surveys at Statistics Canada (Laflamme and St-Jean, JSM 2011).  Three distinct phases during data collection Early in collection – attempt all cases Mid collection – increase response rates Late collection – reduce variability of response rates between domains of interest  Key to success is changing from one phase to the next at the optimal time 25/10/2015 Statistics Canada Statistique Canada 8

Responsive Collection Design (RCD)  The turning point decisions are based on the comparison of quality indicators to pre-set target levels  Indicators are derived from paradata from current and previous collection activity If targets are too high or too low the turning points will not be effective at improving quality Targets need to reflect the priorities, for example to reduce costs, improve response rates, or optimize both simultaneously 25/10/2015 Statistics Canada Statistique Canada 9

Selective Editing and Top-Down Approach  Data editing is a quality assurance activity, not a data correction activity (John Kovar, 199?)  Goals: Make data fit for use (not perfect) - effectiveness Use as few resources as necessary - efficiency  Human resources to do telephone follow-up calls and manual analysis and data modification are costly  Managers need a mechanism to improve efficiency and effectiveness of manual processes without significantly impacting accuracy 25/10/2015 Statistics Canada Statistique Canada 10

Selective Editing and Top-Down Approach  Focus effort where it will do the most good (Hedlin, UNECE 2008)  Tackle efficiency and effectiveness from two angles: Choose certain units or domains of units for further processing, cease processing of the rest Arrange the units requiring further processing in priority order  Pro-actively control the impact on quality by basing turning points and priorities on comparisons of quality indicators to pre-set targets 25/10/2015 Statistics Canada Statistique Canada 11

Selective Editing  When to stop processing Quality indicator could be mean squared error, coefficient of variation, response rate, calculated for key variables at cell or domain level  Targets need to be set carefully If too high, might never be reached, end of processing will never be triggered, costs will not be reduced If too low, resulting data might not be fit for use 25/10/2015 Statistics Canada Statistique Canada 12

Top-Down Approach  How to prioritize units needing more processing Score function – to get a single rank incorporating several different criteria simultaneously “Biggest” based on some size measure (prior knowledge) “Biggest” based on a measure of impact (relative to what has already been collected) Most outrageous errors (outlier detection) 25/10/2015 Statistics Canada Statistique Canada 13

An Example: Edit process  Canadian Census of Population edit and imputation process  110 modules grouped into 43 processes  Underwent a “Quality Assurance Review” in 2013 (Reedman and Julien, FCSM 2013).  65-70% of time was spent on manual data verification  Recommended increased use of automation, and pre-set quality thresholds to limit activity that amounts to “polishing the apple” 25/10/2015 Statistics Canada Statistique Canada 14

An Example: Edit process  Pro-active quality management could include: Derive quality indicators for key variables, compare to pre-set targets, and direct satisfactory records onwards to the next processing step, while only retaining unsatisfactory records for appropriate intervention Use a top-down prioritization method to further restrict manual intervention to only records having a significant impact  The effect on data accuracy and potential time (cost) savings could be estimated using Census 2011 data 25/10/2015 Statistics Canada Statistique Canada 15

Conclusions  Many sources of error in statistical processes  We looked at four ways to manage accuracy and timeliness in data collection and manual processing Interviewer monitoring Responsive Collection Design Selective editing Top-down prioritization Using paradata Feasibility and effectiveness demonstrated Can be used separately or together 25/10/2015 Statistics Canada Statistique Canada 16

Thank-you! For more information, please contact: Laurie Reedman Statistics Canada 25/10/2015 Statistics Canada Statistique Canada 17