South Africa - Measuring Social Connection and Children’s Well-Being Using Multiple Data Sources
Reference ID | AHPU.ZA011.SOCIALCONN.2016 |
Year | 1992 - 2007 |
Country | South Africa |
Producer(s) |
Randall Kuhn Mark Collinson Sangeetha Madhavan Nicholas Townsend |
Metadata | Documentation in PDF |
Created on
May 03, 2018
Last modified
May 21, 2018
Page views
51373
Sampling
Sampling Procedure
We plan to use AHDSS data elements to build roster of kin by relationship type (mother, father, spouse, child, sibling) and populate these rosters with key network characteristics (age, location, education, survivorship). Having built the database, we will engage in preliminary models of network coverage, in comparison to total network size and to the relevant kin identified in CWSC’s typology of relevant kinship impacts. Coverage will be estimated by age of primary individual, by kinship type, and over time. We will conduct bivariate and multivariate assessment of the representativeness of the kinship database by comparing background differentials between individuals having full network coverage and those with more limited coverage. Finally, we will begin to conduct multivariate regression models of changes in child outcome status using 1997 and 2003 as baseline and 2003 and 2006 as followup periods, and event history models of child survivorship using 1997 and 2003 as baselines and all subsequent vital events as a followup period.