Understanding Society User Support

Dear Stephanie
thank you for your quick reply. However, I have already tried to change Alina's code by adding b before `w' (and I have already put all the files into one folder), but when I run the do file I have a problem with the variable "bm_birthy". In particular, the commands generate a variable "bm_birthy" with all missing values. So i end up having no information on the year of birth of the mother.

For how it concern hiqual_dv I don't understand how to derive the vocational background from this variable. If I am not wrong hiqual_dv is a categorical variable with the following categories:
Degree
-Other higher
-A level etc
GSCE etc
- Other qual
-No qual

which does not tell me much about the vocational background. Ideally, I would like to derive a variable like qfvoc or isced in the HBPS.

Any suggestions will be very much appreciated.

Thank you,
Maria

Dear Maria,

We have w_qfvoc1-16 in Understanding Society, which could be used to derive a variable similar to bw_qfvoc, but please check the question wording, universe and available responses are close enough for your purposes.

I'm not sure what could cause the problem you have with bw_birthy. Does it only happen in Wave 13?

Best wishes,
Stephanie

Dear Stephanie,
thank you for your reply.
I tried to generate a variable from w_qfvoc1-16 and it worked. I was able to create a dummy variable equal to 1 if the individual has a vocational qualification, zero otherwise. Do you have any suggestions on which variable should be compared with this new generated variable in order to generate another variable that takes value one whether the HIGHEST qualification achieve by the individual is a vocational qualification rather than academic?

For how it concerns the year of birth for children, I don't think I can use the variable mnpid for the Harmonized BHPS since only mnpid_bh is available in indall.dta. However, while mnpid in the Understanding Society is the natural mother pidp, mnpid_bh (HBHPS) is the natural mother pid. The latter should be the personal identifier, while the pidp is the cross-wave identifier. I think this is the reason why Alina's code is not working with the harmonized version (My problem with bw_birthy was for all the waves).
So probably instead of using pidp w_birthy and mnpid as for the Understanding Society, for the Harmonized BHPS i should use pid w_birthy and mnpid_bh, doing the following (I tried to keep even the pidp variable so that I would be able to match the new dta with indiresp.dta.):

foreach w in a b c d e f g h i j k l m n o p q r {
use pidp pid b`w'_birthy b`w'_mnpid_bh using b`w'_indall, clear
save temp, replace
use temp, clear
keep pid b`w'_mnpid_bh
keep if b`w'_mnpid_bh>0 & b`w'_mnpid_bh<.
rename pid kpid
rename b`w'_mnpid_bh pid

merge m:1 pid using temp, keepus(b`w'_birthy pidp)
    drop if _m==2
    drop _m

rename b`w'_birthy b`w'_mbirthy
    keep pid kpid b`w'_mbirthy
    rename pid b`w'_mnpid_bh
    rename kpid pid
    merge 1:1 pid using temp
    drop if _m==2
    drop _m

rename b`w'_birthy b`w'_kbirthy
    rename pid kpid
    rename b`w'_mnpid_bh pid
    save b`w'_mother_child, replace
}

Any comments are more than appreciated.

Thank you very much

Maria