BHPS and UKHLS weights
For our project we are working with wave 18 of BHPS and wave 1 of UKHLS. In these datasets we are using cross-sectional and design weights, respectively. At first we have simply appended the weights, but we've just noticed that the two have drastically different support. Cross-sectional weight from wave 18 ranges from 0 to 5.3, while design weight from wave 1 ranges from 0 to 6015.3. Could you please help us understand the difference and suggest ways to make them on a similar scale?
#1 Updated by Olena Kaminska 11 months ago
You shouldn't use design weights on their own - they are provided for advanced users who want to create their own nonresponse adjustment.
I suggest that you use an equivalent weight from UKHLS wave 1 to the one that you use in BHPS dataset.