2010 5 10 Post Enumeration Survey for Population

2010. 5. 10 Post Enumeration Survey for Population Census Jaewon Lee Statistical Research Institute Statistics Korea #/24 /22

Outline of the 2005 PES § PES has been conducted on every 5 years § PES period : 12. 1. ~ 8. (8 days) - Census period : November 1. ~ 15. § Sample size : 730 enumeration areas - 0. 27% of total enumeration areas § No. of questions : 15 questions § Data collection method : Face to face interview 2/12

Objectives of PES § to estimate coverage errors of the population census § to evaluate the quality of answers by measuring levels of agreements between census and PES answers § to improve planning and operations of the census by characterizing missed and erroneously included individuals § to produce information for population, households projection 3/12

Organizational structure PES Planning Census Vital Statistics Div. Population Census Div. * Trends Analysis Div. in the Statistical Research Institute (2010) Field Management Local Office Local Government 4/12

Recruiting for enumerators § Enumerators of PES : Well- trained field staff in the local office of Statistics Korea (730 people) Ø Enumerators of the Census : temporary hired people during census-taking periods (105, 000 people) § Benefits of using field staff in the local office Ø Extensive experience in field survey Ø Produce high quality survey data § Criteria for selection : experience in data collection for household survey 5/12

Sampling design § Sample size : 0. 27% (730 enumeration areas) Ø 60 households per one enumeration area Ø Sample size of provinces were allocated using standard errors in 2000 PES and compromise allocation methods § Sampling frame : list of census enumeration areas § Sampling methods Ø Stratified by the type of enumeration area Ø Proportionate probability sampling within a stratified list of census enumeration area § Analysis domains : national and province 6/12

Field Work § Census Dependent re-enumeration (single system) Ø Information provided to PES enumerators ü Census map ü Household list ü Basic information for household members : name, gender, age, marital status, relationship to the head of the household Ø Enumerators duties ü Identify/probe all households in an EA boundary ü Face to face interview ü Persuade interview refusals 7/12

Questions for § Question for individuals PES Ø Name, gender, age, marital status, relationship to the head of the households Ø Date moved to the current address, type of individuals (nonmover, in-mover, out-mover, erroneously included, omitted), reason for omission, reason for erroneous inclusion, previous address of in-mover § Question for households Ø Types of household, reason for omission, reason for erroneous inclusion, house types 8/12

Coverage errors in PES 1995 2000 2005 Net coverage error (%) 1. 25 1. 56 0. 90 Omission rates(%) 3. 06 3. 30 2. 39 Erroneous inclusion rates (%) 1. 81 1. 74 1. 49 Gross coverage error 4. 87 5. 04 3. 88 9/12

Use of PES results § Adjust the base population of the population projection by coverage errors § Adjust the base household numbers for the households projection § Improve future census planning and operations § Evaluate census coverage errors § Evaluate quality of answers § Census figures were not adjusted by PES results 10/12

General directions for § Expand sample size from 730 EAs to 3, 000 EAs 2010 PES § Estimate coverage errors to the city and county level (nation, province, city/county) § Intensify independence of PES from the census § Produce information to evaluate the registered population 11/12

Thank You 12/12