Consultation on the design of the Scottish Household Survey (SHS) for 2017 and beyond

Footnotes

1. Annex A and the 2015 SHS questionnaire can be used to look at exactly what questions are protected under each of the sections. Instructions on how to do this are included in Annex A.

2. Under option B, current one third sample questions become half of the new sample size, rather than one third, and therefore would need to feature in 6 streams of the survey rather than 4 streams. Due to the use of streaming of questions, this also means the sample size needs to increase slightly from 3550 (one third of 10700) to 3750 (one half of 7500).

3. When 2017 data becomes available, it would be combined with 2016 data to produce two year rolling average estimates, which would be published in summer 2018. Assuming this option is continued for SHS 2018, when 2018 data became available it would be combined with 2017 data to produce the next two year rolling average estimates which would be published in summer 2019.

4. The smallest local authorities currently have a target annual sample size of 250. Under option B, assuming a proportionate around 30% drop across all local authorities, this would decrease to 175 on an annual basis, giving a combined sample size of 350 over two years. The confidence interval for a sample size of 250 on point estimates of 70% and 30% (yes/no question) is +/- 6.5% compared to +/- 5.5% for a sample size of 350. The Scottish Government will review options for the optimal allocation of the overall sample across local authorities whilst still maintaining a representative national sample.

5. The SHS Steering Group consists of internal (Scottish Government) and external members including local authority and Improvement Service representation.