1. Like many national surveys of households or individuals, in order to attain the optimum balance between sample efficiency and fieldwork efficiency the sample was clustered. The first stage of sampling involved randomly selecting postcode sectors. The sample frame of postcode sectors was also stratified (by urban-rural, region and the percentage of people in non-manual occupations) to improve the match between the sample profile and that of the Scottish population. For further details of the sample design, see Paragraph 5.1.
3. Pre-calibration weights account for differential selection probabilities (e.g. oversampling of rural areas), or non-response on an area level. When applying these weights we are left with possible bias due to non-response on an individual level.
4. The design effect, often called just deff, quantifies the extent to which the expected sampling error in a survey departs from the sampling error that can be expected under simple random sampling. The design effect increases for more complex sample designs and when weighting adjustments are applied to the final results of the survey.
5. See http://www.gov.scot/Topics/Statistics/SIMD/ for further details on the SIMD.
6. These variables were created by the ScotCen/NatCen Survey Methods Unit. They are based on SIMD scores for all datazones, not just those included in the sample - so an individual who lives in the most deprived quintile of Scotland will also be included in the most deprived quintile in the SSA dataset.
Email: Donna Easterlow