Cross-sectional weight for self-completion questionnaire in wave 6
I've been working with the 'satisfaction' questions from the adult self-completion questionnaire, looking at these in cross section at each wave. For waves 2-5 I've been using the cross-sectional weight w_indscub_xw but an equivalent is not available for wave 6. The only cross-sectional weights I can see available are f_indpxui_xw and f_indinui_xw for the adult/proxy or adult main only interviews.
Why is there no f_indscub_xw variable for wave 6, and which of the available weights should I use for cross-sectional analysis of the self-completion items?
#1 Updated by Victoria Nolan over 3 years ago
- Status changed from New to In Progress
- Assignee set to Mark Fransham
- % Done changed from 0 to 10
- Private changed from Yes to No
Many thanks for your enquiry. This has been passed on to our weighting team who will get back to you shortly.
Best wishes, Victoria
On behalf of the Understanding Society Data User Support Team
#2 Updated by Peter Lynn over 3 years ago
- File f_indresp_in_sc_px_xw.dta f_indresp_in_sc_px_xw.dta added
- Category set to Weights
- Target version set to M6
- % Done changed from 10 to 90
We decided to discontinue indscub_xw, but following various enquiries from users we have subsequently produced this weight (which is for analysis of items that were not asked of the new IEMB sample at wave 6), and f_indscui_xw (for items that were asked of the whole sample). These additional weights will be included in the next update to the data, but are not currently available in the released data.
The items f_sclfsat1 f_sclfsat2 f_sclfsat7 and f_sclfsato were not asked of the IEMB, so f_indscub_xw is the weight you want.
Please find this in the attached file, which can simply be merged with f_indresp.
#3 Updated by Mark Fransham over 3 years ago
Thanks for the quick response Peter - much appreciated. I've downloaded the data and 4,671 of the rows in the weights file are missing (ie empty cells, not just coded as missing). Is that right? I would have expected all rows to have valid values for the weights.
Also, out of interest - why did you decide to discontinue indscub_xw?
#4 Updated by Peter Lynn over 3 years ago
Most of these (4,656) are the IEMB sample, who by definition do not have any of the "ub" weights. The other 15 are odd cases that cannot be assigned a weight due to missing crucial household grid information. These should be treated as having a weight of zero for each of the 3 ub_xw weights.
The decision to discontinue was due to a need to rationalise. We had considerable additional weighting work to do for the wave 6 data release, due to the IEMB sample, so we had to drop some tasks in order to be able to hit the release deadlines with our limited resources. We chose production of indscub as one of the tasks to drop, as response rate to the self-completion (conditional on giving the individual interview) had been rising over the waves, and was 95% at wave 5 without strong demographic differences. We therefore concluded that it would be adequate to use indinub to analyse self-completion items.