Support #1168
open
Selecting correct weighting
Added by Ningqian Yang almost 6 years ago.
Updated almost 4 years ago.
Description
Hi,
We are now conducting a housing research regarding the following three questions:
1. Comparison of the proportion of PRS Vs social housing tenants Vs owner occupiers who have problems paying for housing; for people in the PRS, compare proportion of households with and without children who have problems paying for housing
2. Comparison of educational aspiration between children living in PRS Vs social housing Vs owner occupied
3. Comparison of health and mental health outcomes between children living in PRS Vs social housing Vs owner occupied
To answer the questions, we will use variables in three data files from Wave 8, which are h_hhresp, h_child, and h_youth. Based on our understanding on guideline in selecting correct weighting, we should use n_ythscui_xw weighting across the whole analysis. We would like just to confirm with you whether it is a correct decision to make. We appreciate expertise in using the dataset and thanks for your kind help in advance.
Many thanks.
Ningqian
- Category set to Weights
- Assignee set to Olena Kaminska
- Private changed from Yes to No
Dear Ningqian,
Thank you for your question. I don't have enough details about your analysis, but it sounds to be that to answer question 1 you would use hhresp and indall datasets. If that's the case you can use n_indenui_xw weight. But if you always include information from youth file in your analysis you would need n_ythscui_xw. Note your analysis would be limited to 10-15 year olds only.
Hope this helps,
Olena
Dear Olena,
Thanks for your kind answer.
Sorry for not having provided enough information in my questions. For question 1, we will use h-hhresp dataset; for question 2, we will use h-youth and h-hhresp; for question 3, we will use h-child, h_youth, and hhresp. As these are three questions in one research, I am not sure whether we should use the same weight for answering all questions, or choose different weight depending on which dataset we used on each analysis. If it is in the first case, we may use n_ythscui_xw throughout the whole analysis;if it is in the second case, I suppose we should use n_hhdenub_xw in question 1 and n_ythscui_xw in question 2 and 3. Is it correct?
Thank you very much for your kind help!
Best regards,
Ningqian
- Status changed from New to In Progress
- % Done changed from 0 to 50
Dear Ningqian,
Thank you for your clarification. Your weight will depend on the datasets you use. When you use hhresp only please use h_hhdenui_xw weight. When you use youth and hhresp please use h_ythscui_xw weight. When you use child, youth hand hhresp datasets please use h_psnenui_xw weight from indall dataset - this is a suboptimal weight which is closest to the weight that you would need for such analysis.
I hope this helps,
Olena
Dear Olena,
Thanks a lot for your answer.
May I ask what if I only use hhresp and h-child in the analysis? Which weight should I use? Is the h_psnenui_xw weight? But I don't think I need variables from indall dataset.
Thanks.
Ningqian
Ningqian,
Yes, the h_psnenui_xw will be the correct weight for the combination of child and hhresp datasets.
Thanks,
Olena
- Status changed from In Progress to Resolved
- % Done changed from 50 to 100
It was really giving me a good idea of doing some research work of mine. will connect if help needed.
- Assignee deleted (
Olena Kaminska)
Also available in: Atom
PDF
Updated by Stephanie Auty 
Updated by 
Updated by 
Updated by 
Updated by 
Updated by