Have a question about using weights in Understanding Society?

We have resources to help you use weights correctly with the Understanding Society dataset. 

Here at Understanding Society we've noticed an increase in the number of questions from data users about choosing and applying weights when analysing the Study. To help researchers with this area of data analysis our weighting team has put together guidance on how to use weights. We've also updated our weighting frequently asked questions and there's a webinar for new users to introduce weights and how to use them. 

Information about weighting

Understanding Society is designed to be used with weights. This guide introduces users to the basics of using weights

Weighting Frequently Asked Questions

You can find our weighting FAQs here.

Our FAQ document covers the questions: 

  • Which population does Understanding Society represent?
  • Can I represent a subpopulation?
  • What do a I need to do to represent a population or subpopulation? 
  • Are sample sizes adequate to represent ethnic minorities or immigrants?
  • Why should I use weights in my analysis? 
  • Which weight should I use for my analysis?
  • What happens if I don't use a weight?
  • Will it be sufficient to include a weight variable in my regression model as a control variable?
  • What happens if I don't correct for clustering? 
  • What happens if I don't correct for stratified sampling?
  • Can I run analysis on a calender year/month?
  • Can I pool data from different waves for cross-sectional analysis?
  • Should I worrty that some members of my analysis sample have a weight of zero?
  • My analysis sample is a subset of one for which weights are provided. What weight should I use?
  • There isn't a weight for the combination of waves and instruments that defines my analysis sample. How do I derive my own? 

Webinar: Weights in Understanding Society

If you have a specific question about using weights please contact the User Support team by email or post your question on the User Forum