i_hidp and hidp missing from jk datasets?
I realised that there are only the j_hidp k_hidp and jk_hidp variables available in the new "mainstage_data_2019" release (jk_indresp_cv & jk_hhresp_cv). Does this mean that these are new households that cannot be linked back to previous waves using the original hidp or i_hidp (Wave 9) variables?
Updated by Alita Nandi 3 months ago
- Status changed from New to Feedback
- Assignee changed from Gundi Knies to Theocharis Kromydas
- % Done changed from 0 to 50
- Private changed from Yes to No
Each person in the Covid19 survey will have been interviewed in 2019 either as part of Wave 10 or Wave 11. This datafile includes any information collected from the Covid19 survey participants during 2019 as part of Wave 10 or 11 interview, whichever was their 2019 interview.
jk_hidp is a combined version of j_hidp and k_hidp. The variable jk_wave shows which wave the data is taken from.
You can link the covid19 survey data files (cW_indresp_w or cW_indresp_t or xsample...) with any of the previous wave (Waves 1-9 or BHPS Waves 1-18) individual datafiles using pidp, and to Wave 9-10-11 household level files using i_hidp, j_hidp, k_hidp.
Understanding Society User Support
Updated by Theocharis Kromydas 3 months ago
Many thanks for your reply.
There are a couple of following up points I would need to clarify with you if I may. Please see below:
"_Each person in the Covid19 survey will have been interviewed in 2019 either as part of Wave 10 or Wave 11. This datafile includes any information collected from the Covid19 survey participants during 2019 as part of Wave 10 or 11 interview, whichever was their 2019 interview_."
It seems that the data is not only from 2019. I have noticed that there is an overlap with covid waves as in the jk_indresp dataset there are individuals whose interview end date is April and May 2020. How do we need to deal with this?
"_You can link the covid19 survey data files (cW_indresp_w or cW_indresp_t or xsample...) with any of the previous wave (Waves 1-9 or BHPS Waves 1-18) individual datafiles using pidp, and to Wave 9-10-11 household level files using i_hidp, j_hidp, k_hidp._"
What I noticed here is that there are no individual observations where i_hidp=j_hidp=k_hidp, not even where j_hidp=k_hidp. So it seems that there are zero individuals who lived in the same household in Wave 9, Wave 10 and Wave 11, or even between Wave 10 and Wave 11, which I find quite strange. Could you please clarify if this is the case?
Updated by Alita Nandi 3 months ago
1. About interviews in 2020: Sorry for not being clearer in my response. The Understanding sample can be considered to be a collection of 8 quarterly fieldwork samples - so quarter 1 is interviewed in the first three months of the 24 month fieldwork period and so on. However, due to respondents being away or busy sometimes it is not possible to interview in their alloted fieldwork period, in which case attempts are made to interview them in the following months. So, Quarter 5-8 samples were expected to be interviewed for Wave 10, and Quarters 1-4 samples were expected to be interviewed for Wave 11. Which meant that Q3 & Q4 (W11), Q7 & Q8 (W10) may have spilled over into 2020. To check this do the following (this is the code in Stata):
tab jk_wave jk_quarter if jk_istrtdaty==2020
2. Why are there no cases with j_hidp = k_hidp that is because household identifiers are not stable across waves. Given that people move out and into different households, it is not possible to aprori determine what is a stable household. So, the household identifier uniquely identifies a household in a wave, but not across waves. If you want to match household information such as household income to an individual across waves, first match the household level information to the individual in each wave and then merge or append (as appropriate) these individual level datafiles.
Hope this helps