dependent interviewing on ons household surveys
DESCRIPTION
Dependent interviewing on ONS household surveys. Rob Bumpstead Office For National Statistics. Surveys covered in presentation. Labour Force Survey (LFS) Large, well established UK household survey Using computer-assisted dependent interviewing since 1991 Continuous Population Survey (CPS) - PowerPoint PPT PresentationTRANSCRIPT
Dependent interviewing on ONS household surveys
Rob Bumpstead Office For National Statistics
Surveys covered in presentation
• Labour Force Survey (LFS)– Large, well established UK household survey– Using computer-assisted dependent
interviewing since 1991
• Continuous Population Survey (CPS)– Proposed new integrated survey, incorporating
LFS and other surveys using dependent interviewing
– Provisional start date January 2008
Labour Force Survey family
• Labour Force Survey (LFS)– Quarterly panel survey– Rotating panel design of 5 waves– C. 55,000 households per quarter
• Annual Local Area Labour Force Survey (ALALFS)
– Annual panel survey– Rotating panel design of 4 waves– C. 90,000 households per annum
How we use dependent interviewing - a variety of approaches
• QLFS– All previous responses ‘fed forward’– Majority questions proactive DI
• e.g Household roster, qualifications and employment history, travel to work, occupation and industry, earnings
– Recent time-bounded questions use reactive DI• e.g Current employment, job search, training
• ALALFS– Smaller portion of responses fed-forward
• e.g Household roster, qualifications, occupation and industry,
Why we use dependent interviewing - pros and cons
• Reduces overall measurement error by • reducing false rates of change• and misdating of status changes
• Improves interview efficiency • Reduces respondent burden
• Some increase in measurement error by• suppressing true rates of change and • encouraging cognitive satisficing
• More problematic with proxy respondents• Greater resource demands
Aims of the CPS development project
• To develop a world class modular social survey system as an integral part of the ONS modernisation programme by January 2008
• To deliver better quality, more reliable estimates at national, regional and sub-regional levels
• To produce more outputs more regularly and coherently on which government and stakeholders can better base decisions
ONS surveys for integration
• Labour Force Survey (LFS)– including annual boosts (ALALFS)
• Annual Population Survey (APS)• General Household Survey (GHS)• Expenditure & Food Survey (EFS)• Omnibus Survey (OMN)
• Scope for other surveys to join CPS after 2008
CPS main features
• A common framework for the 5 continuous household surveys on which ONS leads
• single modular survey instrument• single sample of addresses with different combinations of
modules allocated to each address• single integrated field force with interviewers able to
administer all types of interview combinations
• Formal consultation document and other information available on ONS web-site
CPS Instrument design
• A common core module of questions administered to all respondents in sample
• Topic modules of questions administered to portions of respondents in sample
• Core and topic modules combined to construct a number of different viable interview combinations
Fixed core Rotatingcore
Detailed labour market
Job search
Job history
Employment pattern
Detailedqualifications and
training
Accident
Sickness
Disability
Childcare
Ad-hocmodules
forhouse-holds
Healthand
use ofhealth
services
SmokingDrinking
HousingAd-hoc
modules forindividuals
Familyformation
Detailed income PovertyLiving
Conditions
Expenditure
Sam
ple
si z
e (
265 ,
000
ho
use
ho
l ds
)
Interview1
Interview length
Interview2
Interview3
Interview4
Interview5
• Standardise approach across CPS interview combinations and other ONS household surveys
• Development of more scripted proactive dependent interviewing instruments
• Development of more automated reactive dependent interviewing instruments
Some issues in dependent interviewing under consideration at ONS
Dependent interviewing on ONS household surveys
Any Questions ?
www.statistics.gov.uk/cps