Support #1899
open
Added by Imogen Farthing over 1 year ago.
Updated about 1 year ago.
Description
Hello.
I am using individual data which I have joined onto household data, by hidp for each wave then put all of the waves into a dataframe (g to l). Then for each household (and wave), I have aggregated up some of the individual responses (e.g. max personal income, average financial security response etc). So I have a dataframe which has, for each household and wave, some household responses and some new columns which I have created. I am hoping to aggregate these up by region and year (e.g. in London in 2019 the average household income was x, and the average financial security answer by household was y) so need to use weights - however I am unsure which weights to use as I'm doing a longitudinal study on households (I plan to use the xhhrel file to link the households between waves).
Any advice would be much appreciated.
Imogen,
Thank you for your question. How do you define household longitudinally? What happens if someone moves out, if household splits or if someone moves in, stays for 2 years, and moves out? Generally, because of this, we don't have any users using household longitudinally, and therefore we don't provide longitudinal household weights. Yet, if you think you have a clear definition, please let me know and I will help you with the weights.
There are two alternatives: to study households cross-sectionally, and to study individuals longitudinally, but you could study their household characteristics.
Hope this helps,
Olena
Olena,
I'm wanting to see how the characteristics of these households (e.g. max personal income, average financial security score) change over time and how that compares to another national dataset over time. So, it might make more sense to look at the household cross-sections in each wave and then compare those on a year by year basis if that makes sense?
So I could aggregate up the average household income for one wave using the cross sectional weights and then I can compare that cross-sectional picture between years?
Let me know if that doesn't make sense.
Imogen
- Status changed from New to In Progress
- % Done changed from 0 to 50
- Private changed from Yes to No
Imogen,
Yes, this is definitely a possible option. In comparison over the years it is best to do it at an individual level with longitudinal weights. If you do it at household level, make sure you take into account that these are the same people, so possibly doing it through a cross-classified model may be a good option for you.
Hope this helps,
Olena
- Status changed from In Progress to Feedback
- Status changed from Feedback to Resolved
- % Done changed from 50 to 100
Also available in: Atom
PDF
Updated by 
Updated by 
Updated by