www.cls.ioe.ac. uk Dependent Interviewing: Seminar, University of Essex 16-17 September 2004 Peter Shepherd Centre for Longitudinal Studies, Institute of Education, University of London Birth Cohort Studies National Child Development Study, 1970 British Cohort Study, Millennium Cohort Study
22
Embed
Www.cls.ioe.ac.uk Dependent Interviewing: Seminar, University of Essex 16-17 September 2004 Peter Shepherd Centre for Longitudinal Studies, Institute of.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
www.cls.ioe.ac.uk
Dependent Interviewing:
Seminar, University of Essex 16-17 September 2004
Peter Shepherd
Centre for Longitudinal Studies, Institute of Education, University of London
Birth Cohort StudiesNational Child Development Study, 1970 British Cohort Study, Millennium Cohort Study
www.cls.ioe.ac.uk
Birth Cohort StudiesNational Survey of Health and Development (NSHD)
Those living in GB born in one week in 1946
National Survey of Health and Development (NSHD)Those living in GB born in one week in 1946
National Child Development Study (NCDS)All those living in GB born in one week in 1958
National Child Development Study (NCDS)All those living in GB born in one week in 1958
1970 British Cohort Study (BCS70)All those living in GB born in one week in 1970
1970 British Cohort Study (BCS70)All those living in GB born in one week in 1970
Millennium Cohort Study (MCS)All those born in selected areas of UK over 12 months
beginning September 2000 in England and Wales, and December 2000 in Scotland and Northern Ireland
Millennium Cohort Study (MCS)All those born in selected areas of UK over 12 months
beginning September 2000 in England and Wales, and December 2000 in Scotland and Northern Ireland
www.cls.ioe.ac.uk
Birth Cohort Studies – Main surveys
Notes1 Initial survey carried out at c8 weeks2 Initial survey carried out at c9 months(Age) Sample surveys
Age of cohort members at time of main surveysAge of cohort members at time of main surveys
www.cls.ioe.ac.uk
Goals of Birth Cohort Studies Modelling causal processes from birth to adulthood leading to current
outcomes and assessing the risk of future outcomes.
Assessing the stability of hypothesised causal processes across cohorts.
Comparing the prevalence of behaviour and attributes across cohorts, ages and periods.
Assessing inter-generational continuities and discontinuities in circumstances, behaviour and attributes.
www.cls.ioe.ac.uk
Why cohort studies are important
Expensive compared to cross-sectional surveys, but are important because:
They tell the linked stories of the lives of the members They record how long someone occupies a given
state They link events across the life course & across
domains They enable investigation of the cause & effects, early
experience to later outcomes
www.cls.ioe.ac.uk
Life course Perspective
Holistic
Transitions & pathways
Interconnectedness
Linked lives
www.cls.ioe.ac.uk
Design Principles Continuity & comparability
Age, cohort & period effects
Spatial effects
Consultation
Harmonisation
Life course perspective
www.cls.ioe.ac.uk
MCS - Design Features Cohort born over 12 month period
Season of birth effects Spread workload of professional interviewers Sampling necessitated
Geographically clustered by electoral ward
Wards disproportionately stratified - 3 types:
advantaged; disadvantaged1; and high minority ethnic2
Content multi-purpose & multidisciplinary
1 Poorest 25% of wards on Child Poverty Index.2 At least 30% of 1991 Census ward population = 'Black‘/'Asian‘ – England only.
www.cls.ioe.ac.uk
MCS – Sweeps & sources of informationMCS1 MCS2
(2001/2) (2003/4)9 months 3 years
Mother Mother
Father Father
Child
Older siblings
Birth records
Medical records
18,819 ?
MCS3 MCS4(2005/6) (2007/8)5 years 7 years
Mother Mother
Father Father
Child Child
Older siblings
Older siblings
Teachers
Education records
Education records
Medical records
Medical records
? ?
Includes c500 ‘new families’ missed by DWP at MCS1
Includes c500 ‘new families’ missed by DWP at MCS1
www.cls.ioe.ac.uk
MCS1 – Achieved samples
ChildrenFamilies
interviewed Partners
Single Parents
Total UK 398 20,646 18,819 18,553 13,599 3,194
ENGLAND 200 13,146 11,695 11,533 8,558 1,853
WALES 73 3,000 2,799 2,761 1,957 590
N IRELAND 63 2,000 1,955 1,923 1,326 376
SCOTLAND 62 2,500 2,370 2,336 1,758 375
Notes ** all productive contacts* counting 'superwards' as a single
Achieved Responses **Number of sample
'wards' *
Target sample as boosted
NB: Wards vary in births/12 months (4-600). To minimise fieldwork problems, small wards combined as 'superwards' with at least 24 expected births/12 months.
www.cls.ioe.ac.uk
MCS1 - Response RatesAchieved Response Rate
In-scope Fieldwork Response Rate
WALES 72% 84%
ENGLAND 68% 82%
SCOTLAND 70% 85%
NORTHERN IRELAND
63% 79%
www.cls.ioe.ac.uk
MCS1 & 2 - Main/Partner Topics
Topic Main PartnerHousehold A: Non-resident parentsB: Father’s involvement with baby C: Pregnancy, Labour and Delivery D: Baby’s health and development E: Childcare F: Grandparents and Friends G: Parent’s health H: Self-completion (Attitudes to marriage, parenting, work, etc)
J: Employment and Education K: Housing and local area L: Interests and time with baby
Topic Main Partner1 Household 2 Parental situation 3 Parent’s involvement with child 4 Child Health 5 Grandparents and Friends 6 Parent’s health 7 Housing and local area 8 Employment and education 9 Employment history 10 Childcare 11 Other Matters 12 Self-completion
13 Older siblings
(Child behaviour, discipline mental health, relationships, attitudes to parenting, drugs, alcohol, life satisfaction etc)
MCS1 MCS2 (including c500 ‘new families’)
more…
www.cls.ioe.ac.uk
Cohort studies - Why use dependent interviewing
With thanks to: Mathiowetz Nancy A. and McGonagle Katherine A. (2000) An assessment of the current state of dependent interviewing in household surveys Journal of official statistics 16 pp 401-418
Why use dependent interviewing Now FutureDesign a more efficient means of collecting information, eliminating recording redundancies
Reduce respondent burden Reduce measurement error associated with responses to open-ended items, specifically those in which slight variation in response wording results in significant differences with respect to a classification.
Reduce or eliminate seam effects
www.cls.ioe.ac.uk
MCS2 - Feed forward/Dependent interviewingMCS2 MCS1 feed forwardRoutes for 'old' / 'new' families MCS1 participationCorrectly identify MCS child(ren) Name / sex / dob Identify ‘main’ / ’partner’ respondents / their whereabouts
‘Main’ / ’partner’ name
Recall period Interview dateGuidance for interviewer ‘Main’ / ‘partner’ sight / language problemBreastfeeding ‘Still breastfeeding’Grandparents involvement with children / grandchild(ren); separation
‘Main’ / ‘partner’ parents alive / dead / separated at last survey
Basic skills problems ‘Main’ basic skills problemsEmployment history ‘Main’ employment status Childcare arrangements / history Childcare arrangementsOlder siblings Older sibling in household grid
www.cls.ioe.ac.uk
MCS2 – Use of feed forward 1
www.cls.ioe.ac.uk
MCS2 – Use of feed forward 2
www.cls.ioe.ac.uk
MCS2 – Use of feed forward 3
www.cls.ioe.ac.uk
NCDS - Follow-ups & information sourcesP M S N C D S 1 N C D S 2 N C D S 3 N C D S 4 N C D S 5 N C D S 6
Exams – details of public examination results were gathered from schools and colleges in 1978
Exams – details of public examination results were gathered from schools and colleges in 1978
Information gathered includes:HealthBehaviourFamilyEducationEmploymentAttitudes
Information gathered includes:HealthBehaviourFamilyEducationEmploymentAttitudes
Notesa: Target sample - Excludes emigrants, refusals & deaths. Includes immigrants at NCDS1-3.b: Achieved sample - At least on survey instrument partially completedc: Mother - Could be Cohort Member or spouse/partner
Notesa: Achieved sample. NB: Target sample excluded emigrants, refusals & deaths; and included immigrants in 1975 & 1980.
Information gathered includes:HealthBehaviourFamilyEducationEmploymentAttitudes
Information gathered includes:HealthBehaviourFamilyEducationEmploymentAttitudes
BCS70(2004)
34
— Subject
Children (1 in 2 sample)
?
www.cls.ioe.ac.uk
BCS70/NCDS 2004: Feed forward/Dependent interviewingBCS70/NCDS 2004 BCS70/NCDS 2000 feed forwardCorrectly identify BCS70 cohort member Name / sex / dob Recall period Interview dateGuidance for interviewer Sight / hearing / speech / reading / learning /
other problem / children aged 0-13Household grid Household grid - name / age /relationship of
household membersPartnership status / history Partner statusPregnancy history / children / adopted children
Children / adopted children
Current housing / history Date moved into current accommodationMarital status / partnership history Marital status / partnership historyPartner education Age partner left educationEconomic activity / job history Economic activityPartner’s economic activity Partner’s economic activityHeight HeightParents in household / separation / death Parents in household / separation / death