Understanding Society User Support

Libby,

Thank you for your question. You situation is complex and unique and we do not have the weight that will take into account all of the nonresponse in your analysis but I will be able to suggest the most suitable suboptimal weight.

From the description of your analysis I understand that you are using adult full questionnaires of both parents and require the information about youth from household questionnaire and grid but do not require the youth questionnaire to be completed. If that's the case your best suboptimal weight is the longitudinal full adult response weight taken from the second wave of each pair of the waves that you analyse. This weight will take account for most of the non-response and the design of the survey except for the nonresponse related to joint parent response (as opposed to an individual response). If you are using information from proxy interviews you should use the equivalent proxy weight (and the missed nonresponse will be smaller).

If all the information you need can be obtained from the household questionnaire, then our weights cover all nonresponse parts (if both parents live in the same household) and you should use longitudinal psnenXX_lw weight.

Hope this helps,
Olena