Support #878
closed
employment histories
100%
Description
Hello,
I try to reconstruct employment histories in wave 1 to analyse implications on health in the later waves. Since employment histories are collected for three quarters of the sample in wave 5, I am using both a_empstat (for the first quarter of the sample) and e_empstat (for the remaining sample). Some employment histories of the second group (which should be present in e_empstat) are missed because some people drop out between wave 1 and 5. Am I correct? If yes, would be better to use only the first quarter of the sample (which is randomly selected)?
Updated by Stephanie Auty about 7 years ago
- Status changed from New to In Progress
- % Done changed from 0 to 10
- Private changed from Yes to No
Many thanks for your enquiry. The Understanding Society team is looking into it and we will get back to you as soon as we can.
Best wishes,
Stephanie Auty - Understanding Society User Support Office
Updated by Stephanie Auty about 7 years ago
- Category set to Data analysis
- Status changed from In Progress to Feedback
- Assignee set to marco tosi
- % Done changed from 10 to 70
Dear Marco,
The first six months is a representative sample, but does not include the Ethnic Minority Boost sample. There is no unambiguous answer as to whether using only Wave 1 or both Wave 1 and Wave 5 employment histories is better. It depends on your research question and methods.
Best wishes,
Stephanie Auty - Understanding Society User Support Office
Updated by marco tosi about 7 years ago
Dear Stephanie,
My sample includes people aged 65-74 in wave 1 (n=5,371). 3,705 respondents have information about their employment histories (collected in wave 1 or 5). I analysed employment trajectories (Latent Class) and to what extent these trajectories are associated with health indicators in wave 1. My concern is that a large part of this sample includes people who survive/do not drop out between wave 1 and 5 and are in better health conditions compared to those for which employment histories are not available. Are my concerns correct? Are there specific weights to correct for this bias? Would you suggest to use only the first quarter of the sample (a_empstat)?
Thank you for your help
Best wishes
Marco
Updated by Stephanie Auty about 7 years ago
- Category changed from Data analysis to Weights
- Status changed from Feedback to In Progress
- Assignee changed from marco tosi to Peter Lynn
- % Done changed from 70 to 60
Dear Marco,
I'll assign this to Peter for some weighting advice.
Best wishes,
Stephanie Auty - Understanding Society User Support Office
Updated by Stephanie Auty about 7 years ago
- Assignee changed from Peter Lynn to Olena Kaminska
Updated by Olena Kaminska about 7 years ago
Dear Marco,
Your problem is common for any researcher who studies elderly population. So, indeed the drop out between wave 1 and 5 is a combination of nonresponse (some of it due to health issues) and death. Our weights correct for nonresponse, but do not correct for death. This means that when you look at waves 1 and 5 together you are looking at people who survived over these years. This is ok if you make it clear in your paper. And this may be the simplest way to use the data - simply use the longitudinal weight from wave 5.
Theoretically you may use a part of our sample but we don't have ready weights for such analysis and their calculation is complex.
Hope this helps,
Olena
Updated by Stephanie Auty about 7 years ago
- Status changed from In Progress to Feedback
- Assignee changed from Olena Kaminska to marco tosi
Updated by Stephanie Auty almost 7 years ago
- Status changed from Feedback to Resolved
- % Done changed from 60 to 100
Updated by Stephanie Auty almost 7 years ago
- Status changed from Resolved to Closed