Support #2225
open
SF-12 missingness wave 2
50%
Description
Hi,
I noticed there is quite a substantial amount of missingness (-9) on the derived SF12 physical and mental component summaries (sf12pcs_dv and sf12mcs_dv). This is 7.72% vs <1% in other waves.
Looking at the single items, this seems to be due to a substantially higher number of respondents coded as don't know (-1). These are in the hundreds and thousands at wave 2. For example, 2,063 are coded don't know (-1) on b_scsf3b.
Could you shed some light on why this is?
Many thanks and best wishes,
Verena
Updated by Understanding Society User Support Team 12 days ago
- Category set to Data documentation
- Status changed from New to Feedback
- % Done changed from 0 to 50
- Private changed from Yes to No
Hello Verena,
In Wave 2, an unusually high number of "Don't know" responses were recorded for the question items scsf2b, scsf3b, and scsf4b, which resulted in b_sf12mcs_dv and b_sf12pcs_dv being coded as -9 for these cases.
It’s worth noting that in Wave 2, the self-completion survey was conducted using a paper questionnaire, whereas in later waves, it transitioned to Computer-Assisted Self-Interviewing (CASI). These three items appeared as the second part of paired questions in the paper questionnaire, and it seems that interviewers may have inadvertently overlooked them due to their placement.
Additionally, the original data provided by the fieldwork agency also reflected these figures, confirming that the issue might originated during data collection.
I hope this information is helpful.
Best wishes,
Roberto Cavazos
Understanding Society User Support Team