This is a general question. I've found that "inapplicable" is a common answer to many questions in the xwavedat dataset, and I've deduced that "inapplicable" can mean:
a. the question was skipped due to a skip pattern
b. the question was never asked in years when the person participated in the survey
c. the question wasn't part of the survey, depending on interview type
d. the question wasn't part of the survey for a regional sample
e. some combination of the above
Is this correct? Also I've noticed that for every question in this dataset, 30% or higher of the values for any field are inapplicable/missing. Is there a general reason for this? Thank you.
#1 Updated by Redmine Admin almost 7 years ago
- % Done changed from 0 to 50
If this is a question concerning the UKHLS XWAVEDAT, then this file contains the latest valid response to a given item. If no valid responses are found, the missing value code -9 is assigned. To find out the specific reason for missingness in a given wave, it is necessarry to look up the item in the response files of that wave. Generally, -8 would be if the respondent was routed out of the item, -7 if not asked during a proxy interview, -2 for refused, -1 for "don't know", and -9 for all other reasons. We have yet to update the XWAVEDAT for the BHPS entrants to UKHLS across all stable characteristics. This information is available on BHPS's cross-wave file of the same name.
#4 Updated by Redmine Admin almost 7 years ago
- Status changed from New to Closed
- % Done changed from 50 to 100
You are correct that there are several possible reasons for inapplicability in BHPS's xwavedat. As for UKHLS, users should refer to the missing value coding on specific waves to gain more insight.