Project

General

Profile

Support #1354

Linking children to adults in Covid-19 study

Added by jennie parnham 3 months ago. Updated 3 months ago.

Status:
Feedback
Priority:
Urgent
Category:
-
Target version:
-
Start date:
06/09/2020
Due date:
06/11/2020
% Done:

60%

Estimated time:

Description

Hello,

I have been analysing the April wave of the Covid-19 data. I have come across trouble linking the children to adults using the Pidp.

Given all the pidp's on the child-level dataset (ca_schild_w) I can only link around half of the 4559 children to an adult(ca_indresp_w). Can every child on the child-level data be linked to an adult on the ca_indresep_w dataset?

I have tried using all the different versions of the pidp(1pg, m ,f, etc) to link with the adult pidp.

I need to link the children to income, geography, income variables and food insecurity variables. I have managed to link to adults on previous waves, but i also need data from the current wave

Any advice or suggestions would be greatly appreciated

History

#1 Updated by Alita Nandi 3 months ago

  • Private changed from Yes to No
  • % Done changed from 0 to 60
  • Assignee set to jennie parnham
  • Status changed from New to Feedback

Hello,

The file ca_indresp_w is the adult level file where pidp uniquely identifies each row. The information provided by parents/guardians about their children is put into a child level file, ca_schild_w. For further details see Section 11.

In the file cs_schild_w pidp_c uniquely identifies each row. So, if you rename pidp_c to pidp and match it to xwavedat in the main survey (released version which includes data from waves1-9), then you will see that you can match all but 310 children to the main survey data.

In ca_schild_w, the pidp of the person who provided information about the child (pidp_c) is given by pidp_m (if female) and pidp_f (if male). So, if you create a variable, pidp which equals pidp_f or pidp_m (whichever is not -8), and then match to ca_indresp_w using pidp then you will see that all cases in ca_schild_w will match to ca_indresp_w.

If you use Stata, then take a look at this syntax which explains what I have said:
(1)
use ca_schild_w, clear
isid pidp_c
rename pidp_c pidp
merge 1:1 pidp using xwavedat

(2)
use ca_schild_w, clear
generat long pidp=pidp_m
replace pidp=pidp_f if pidp==-8
merge m:1 pidp using ca_indresp_w

Does this answer your question? If not, please let me know.

Best wishes,
Alita

Also available in: Atom PDF