Weighting in BHPS
I have a problem with weighting the data for BHPS wave 11 which I am hoping you can help me with.
We have created a data file which merges the hhresp file with the indresp file by extracting individuals successively by person number, matching the files to the household file and then adding them all together, selected by whether the respondent answers the question of whether they have natural children (variable name KLPRNT). This gives us a sample of 6476.
However, when we use the cross-sectional individual response weight (KXRWGHT) we have a resulting sample size of 158.
Clearly, this is because of the number of cases that are weighted as ‘0’(i.e. missing) – a crosstab of weight by natural parent status gives a total of 280 non-zero cells.
Are we missing something that enables us to use the full set those respondents who answered the natural parent question, weighted by non-response of individuals?
I have tried to be brief in my explanation and hope this is clear, but if more detail is required please let me know.
Kind Regards, and thanks in advance.
Updated by Redmine Admin almost 11 years ago
- Category set to Data analysis
- Status changed from New to In Progress
- Assignee set to Redmine Admin
- Target version set to BHPS
- % Done changed from 0 to 50
BHPS consists of a number of samples including some regional refreshment samples. The weight, wRWTUK1, can be used to adjust results to the UK population from Wave 11 onwards. See details about the different weights and sample populations in the user guide, p. 197 https://www.iser.essex.ac.uk/bhps/documentation/vola/vola.html