Understanding Society User Support

Dear Gabriella,

Many thanks for your enquiry, and apologies for any confusion. I think in that earlier issue, the link simply takes you back to an earlier point in the post, where the following advice was given:

In the BHPS as all relationships within the household were identified with respect to the household reference person (HRP), the HRP indicator variable was computed. In Understanding Society (UKHLS) everyone's relationship is asked with respect to everyone else, so HRP is not needed and hence not computed. If you need that variable you will have to compute it yourself. The definition of HRP is the person who owns or rents the accommodation and if more than one, the eldest.

Both in the BHPS and Understanding Society, the household questionnaire is completed by someone who knows most about household payments like bills, rent etc. In most cases this is the HRP but not always. If you want to know the person who answers the household questionnaire then that information is available in the variables, w_ivh* (in UKHLS) and wivh* (in BHPS).

I hope this (plus the responses to the earlier issue #290) helps? Please don't hesitate to ask if you would like some additional help.

Best wishes, Victoria.

On behalf of the Understanding Society User Support Team.

Hello Gabriella,

You can construct the hrp variable for Understanding Society using the variables:
w_hidp, w_pno, w_hsownd, w_hsowr1*, w_rentp*, w_dvage

But there will be few households for whom you will not be able to construct the variable due to:
(1) house is netiher owned nor rented (w_hsownd = 97)
(2) house owner or renter is someone outside the household (w_hsowr10 =1 or w_rentp0=1)
(3) house owner or renter information is missing (-9)

Once you have identified the HRP, then you can merge that dataset with the w_egoalt file to identify the live in partner of the HRP and their children. To learn more about how to use the w_egoalt files to construct such relationship variables see Example 8 (Using egoalt files) in the online Stata and SPSS courses:
https://www.understandingsociety.ac.uk/documentation/training/online/
If you use SAS then see the "Using Egoalt Files" example in https://www.understandingsociety.ac.uk/documentation/training/online/sas

I am not sure, but it seems you are saying that the wrel variable in wegoalt files in BHPS shows the relationship between the HRP and others in the household. That is not the case. The wrel variable in the BHPS wegoalt file shows the relationship between every pair of household members.

Hope this helps.

Gabriella Melis wrote:

Hi,

Thank you for your answer below.

Hence, for UndSoc we must assume that the person who answered the HH questionnaire is the HRP, whilst in BHPS this was completely clear, right? Or is there a more precise solution, please?

Considering this issue, may you please suggest the best way (and the variable names) to compute for UndSoc a variable that reproduces the same information available in the BHPS EGOALT file for the wrel variable:
-HRP
-their live-in partners
-their offspring (18+ yo)

Thank you very much.

Best wishes,

Gabriella

Garbiella Melis wrote:

thank you very much for your suggestion, although instead of the w_hsownd I have used the derived tenure variable, and I have not used age as it appeared redundant.

The straighforward and extremely helpful BHPS variable on relationships bewtween the HH members and the HRP I was talking about is not the wrel in the EGOALT file, rather the ahgr2r variable in the wINDRESP file. I think this same variable should have been made available for UndSoc too in the w_INDRESP file, both for comparison with BHPS, as well as for its general importance for the analysis of interactions between HH- and individual-level characteristics.

I'd be happy to share my syntax as soon as it is ready, if anyone else needs it.
Thank you.
Best regards,
Gabriella

Thank you Gabriella for offering the code.
I will write to you directly about this.
Best wishes,
Alita