Support #1175

Can the cross-sectional self-completion interview weight variable be used to estimate the real population distribution at the Local Authority level?

Added by Jing Shen over 5 years ago. Updated almost 2 years ago.

Start date:
% Done:



We're currenlty working on a study, for which we need to know the percentage of each religous group in every LA unit. The Census data seem to have information only about England and Wales (no information is avaiable for Norhtern Ireland and Scotland). Also, small reliogus groups do not seem to be found in Census either. So, in each wave, we've used the cross-sectional self-completion interview weights to weigh the number of resondents in the corresponding group and the total number in every LA unit. We then calculate the percentages based on this pair of numbers in every unit. Can those numbers represent the real populaton distribution (by religon) at the LA level?


Updated by Alita Nandi over 5 years ago

  • Status changed from New to In Progress
  • Assignee set to Olena Kaminska

Many thanks for your enquiry. The Understanding Society team is looking into it and we will get back to you as soon as we can.

Best wishes,
Understanding Society User Support Team


Updated by Alita Nandi over 5 years ago

  • Private changed from Yes to No

Updated by Alita Nandi over 5 years ago

  • % Done changed from 0 to 10

Updated by Olena Kaminska over 5 years ago

Dear Jing Shen,

Thank you for your question. This question is rather a statistical one, and I can only answer about Understanding Society estimates.

I understand (though I am not sure my understanding is correct) that you estimate religious groups in each LA area using Understanding Society. This may be possible using our data, but a few statistical points need to be kept in mind: firstly, pay attention to confidence intervals - I expect them to be relatively large as you will be working with relatively small sample sizes (your sample size will that of each LA); also as you are looking at the proportions occurring towards the tail of a distribution (e.g. less than 5%) I would suggest to use appropriate confidence intervals for the extreme proportions - these will be larger and often skewed to one side. Of course your estimates should be weighted.

Hope this helps,


Updated by Stephanie Auty over 5 years ago

  • % Done changed from 10 to 70

Updated by Alita Nandi about 5 years ago

  • Assignee changed from Olena Kaminska to Jing Shen

Updated by Alita Nandi almost 4 years ago

  • Status changed from In Progress to Feedback
  • % Done changed from 70 to 90

Updated by Understanding Society User Support Team almost 2 years ago

  • Status changed from Feedback to Resolved
  • % Done changed from 90 to 100

Also available in: Atom PDF