The future of Scottish Government Cross- Sectional Population Surveys Sara Grainger Survey Methodology and Co-ordination, Office of the Chief Statistician
Jan 14, 2016
The future of Scottish Government Cross-Sectional
Population Surveys
Sara GraingerSurvey Methodology and Co-ordination,
Office of the Chief Statistician
LONG TERM STRATEGY FOR POPULATION
SURVEYS IN SCOTLAND 2009 -2019
1. To ensure that the Scottish Government’s population surveys meet key information needs while maximising the analytical potential of the data they generate, the precision of estimates and value for money.
2. To give full consideration to issues of survey participation, respondent burden, data quality and data security and to make recommendations that align survey practice across Government and promote good practice to other public bodies.
SHS
SHeS
SCJS
Local survey
Ad-hoc survey
n=11,000
n= 6,000
n=13,000
n= 1,000
n= 2,000
core questions
topic specific questions
Combined sample of core questions from all surveys
n=33,000+
Core Questions and Pooled Samples
Core Questions and Pooled Samples
• Household grid (household relationship matrix / household type)
• Gender• Age / Date of birth• Marital status• Ethnicity• Disability or long term
health• Self-assessed health• Smoking• Caring responsibilities• Mental wellbeing
• Sexual identity• Religion / Belief• Educational attainment• Household income• Economic activity• Tenure• Car ownership / access• Country of birth• Crime victimisation and
reporting• Perception of local crime
rate
Scottish Crime and
JusticeSurvey
SHS 2010 Review
• Full “roots and branches” review
• User questionnaire, focus groups & discussion at Survey Conference earlier this year
• Rejuvenated SHS Steering Group with wide representation
• Widely used for variety of purposes
• Competing demands for breadth and depth
• Dissatisfaction at 2 year sample design
• Most variables change little year on year
Options for the future
• Alternative sources • Different modes• Reduce content
– Merge surveys
• Reduce sample sizes– Precision– Sub-population breakdowns – Frequency
Household Health Crime & Justice
House Condition
Health
Crime & Justice
House Condition
Merging Surveys
SHCS questions
follow-up dwelling survey
Reducing Sample Sizes
Household 14,000 11,000?
Health 6,000
Crime & Justice 16,000 13,000every other year
House Condition 3,000 3,000
Implications of reduced sample sizes
1. Reduced precision
0
10
20
30
40
50
60
70
80
90
100
n=500 n=250
Implications of reduced sample sizes
2. Reduced capacity to drill into data sub-national geographies
sections of the population
Whole sample 10,000 5,000
Sample of adults age 60+ 2,400 1,200
Sample in social rented sector 2,200 1,100
Sample of adults age 60+ in the social rented sector
530 265
Implications of reduced sample sizes
3. Reduced frequency of reportingpotentially maintaining precision and capacity to drill into data
by combining 2, 3, 4 years of data together
What are your priorities and why?
1. Precision? Why?
2. Capacity to drill into data? Why?
3. Frequency of reporting? Why?
Would you prefer all 3 to be reduced a bit, or one to be reduced a lot? Why?
What would the implications be for your ability to deliver outcomes?
Further InformationThe Long Term Survey Strategy, papers and minutes from the Population
Surveys Co-ordinating Committeehttp://www.scotland.gov.uk/Topics/Statistics/About/Surveys
Core and harmonised survey questionshttp://www.scotland.gov.uk/Topics/Statistics/About/SurveyHarm
Information about the Scottish and UK surveys, and full topic guide to SG surveys
http://www.scotland.gov.uk/Topics/Statistics/About/SurveyHarm/surveytopiccoverage
Survey Methodology and Co-ordination Branch [email protected] 0131 244 0329
[email protected] 0131 244 3339
To be kept informed select “Population and Household Surveys” in your areas of interest on the ScotStat register