Support #381
lvwhy recording
100%
Description
Hello,
this query is in regards to the recording practices of 'leavers' in variable lvwhy in the BHPS (1991 to 2009 data).
We have found cases where an is individual being recorded as 'deceased' (variable lvwhy), after one or more years of being recorded as 'other', 'don't know' or 'current' - without responding to any survey questions.
Is it possible to get any information on the actual year of death for these individuals, or are there any estimates/rules for what to assume? In other words, is it likely that the individual has indeed been 'non-response' during these years or rather deceased but unreported?
I've consulted the forum issues #216 & #215.
Many thanks in advance!
Updated by Redmine Admin almost 9 years ago
- Category set to Data inconsistency
- Status changed from New to In Progress
- Assignee set to Valentina Zigante
- Target version set to BHPS
We have yet to obtain permissions to link to mortality registers. We would like to replicate your analysis. How many cases did you find?
On behalf of the team,
Jakob
Updated by Valentina Zigante almost 9 years ago
Thank you for the quick response.
I've found 1,253 observations, where the length of non-response prior to being recorded as 'deceased' ranges from 1 to 15 years.
Kind regards,
Valentina
Updated by Redmine Admin almost 9 years ago
- % Done changed from 0 to 50
Is it possible to get any information on the actual year of death for these individuals, or are there any estimates/rules for what to assume? In other words, is it likely that the individual has indeed been 'non-response' during these years or rather deceased but unreported?
No and there is of course less certainty about non-response years/exit. You may find the outcome variabels IVFIO on the cross-wave file XWAVEID easier to work with.
On behalf of the team,
Jakob
Updated by Redmine Admin almost 9 years ago
- Status changed from In Progress to Closed
- % Done changed from 50 to 100