Support #1927
open
Added by C Rowan Marron over 1 year ago.
Updated about 1 year ago.
Category:
Data inconsistency
Description
Dear sir or madam,
The variables SF1 and SCSF1, in the w_indresp files, are supposed to represent self-assessed health on a 5-point scale, and are supposed to be the same. The documentation for these variables says that they should be combined together. However, for waves 2-5 inclusive, there are a large number of cases, about 23% of total data, where both are available and not null, they differ from each other. That is, the same person in the same wave reports their health as 'very good' on one variable and 'excellent' on the other, for example. Please advise how this can happen and which variable should preferentially be used in this situation.
Kind regards,
Rowan.
- Status changed from New to Feedback
- % Done changed from 0 to 80
- Private changed from Yes to No
Hello Rowan,
Thanks for highlighting this. We are very sorry the variable note is not entirely correct. We will revise it.
In Waves 1-5, SF1 was asked of everyone by the interviewer (you can check this in the questionnaire), but from Wave 6 onwards, this was done only for proxy respondents. From Wave 2 onwards the same question was also included in the self-completion part of the individual questionnaire and was renamed SCSF1. So, for Waves 2-5, there are two types of responses available for this question (in variables SF1 & SCSF1) for those F2f respondents who also completed the self-completion part of the questionnaire.
Hope this helps.
(Alita)
Best wishes,
Understanding Society User Support Team
- Status changed from Feedback to Resolved
- % Done changed from 80 to 100
Also available in: Atom
PDF
Updated by