Support #154

'0' weight in BHPS

Added by David Bayliss over 10 years ago. Updated about 10 years ago.

Start date:
% Done:



Good afternoon,

I have a question regarding the weight variable xrwtuk1 in the BHPS. For my analysis I have selected cases who have a valid response to my two main vairables (economic status and GHQ-12), and then had a look at the weights. Within the cross-sectional weight variable for the UK (xrwtuk1) I have around 2.5% of cases which are '0' weighted. All of the cases I have selected are full interviews, and the '0' weighted cases have a range of sample origins (67% are OSMs). I have read the weighting section of the user manual but I cannot find a reason that matches the data. Please could you advise me on what may be the cause of the '0' weight?

Kind regards,
David Bayliss


Updated by Redmine Admin over 10 years ago

  • Status changed from New to In Progress

David, Would like to replicate this - please let us know the exact variable names and wave(s). Jakob


Updated by Redmine Admin over 10 years ago

It looks like the weights could be missing in a small proportion of cases. We would like to look into this in more detail and update you once we have got some results.


Updated by Olena Kaminska over 10 years ago


The reason for 0 values in cross-sectional weights is most likely due to no OSM within a household having longitudinal main interview non-zero weight in the wave you are looking at. Cross-sectional weights are inferred from longitudinal weights, and if every OSM in a household have 0 longitudinal weights, there is no positive value to infer. As you indicated this is a problem for only a small portion of a sample.

Hope this helps,


Updated by Redmine Admin over 10 years ago

  • % Done changed from 0 to 50

Updated by Redmine Admin about 10 years ago

  • Status changed from In Progress to Closed
  • % Done changed from 50 to 100

Also available in: Atom PDF