Ghana - Study on Global Ageing and Adult Health 2007-2008, Wave 1

Ghana used a stratified, multistage cluster design.
The sample was stratified by administrative region (Ashanti, Brong Ahafo, Central, Eastern, Greater Accra, Northern, Upper East, Upper West, Volta and Western) and type of locality (urban/rural) resulting in 20 strata and is nationally representative.
The Census Enumerated Areas (CEA) of the 2000 Population and Housing Census was used as the sampling frame.
A sample of 251 EAs was selected as the primary sampling units (PSU). One of the selected PSUs was not used. This was because the EA which was expected to be located at Korle Bu Teaching hospital cccould not be traced. The number of EAs to be selected from each strata was based on proportional allocation (determined by the number of EAs in each strata specified on the census frame). EAs were then selected from each stratum with probability proportional to size; the measure of size being the
number of individuals aged 50 years or more in the EA.
In each selected EA, a listing of the households was conducted to classify each household into the following mutually exclusive categories:
(1) WHS/SAGE Wave 0 follow-up households with one or more members aged 50 years or more;
(2) New households with one or more members aged 50 years or more;
(3) WHS/SAGE Wave 0 follow-up households which did not include any members aged 50 years or more, but included residents aged 18-49; and,
(4) New households which did not include any members aged 50 years or more, but included residents aged 18-49.

Twenty-four households were randomly selected from each selected EA. All WHS/SAGE Wave 0 follow-up 50-plus households were eligible for the household interview (one household respondent was selected). Twenty such households were selected. If this target number was not reached, then the balance was selected using systematic sampling from the new 50-plus households. All 50+ members of the household were eligible for the individual interview (multiple individual interviews possible in these households).

Stages of selection
Strata: Region, Locality=20
PSU: EAs=235 surveyed
SSU: Households=5269 surveyed
TSU: Individual=5573 surveyed


One of the 251 selected PSUs was not used. This was because the EA which was expected to be located at Korle Bu Teaching hospital cccould not be traced.


Household
Response rate=86%
Cooperation rate=98%

Individual:
Response rate=80%
Cooperation rate=92%


Household weights(variable hhweight) for analysis at household level and individual weights(variable pweight) for analysis at person level were calculated.
These were based on the selection probability at each stage of selection. Household weights were post-stratified by region and locality according to the 2010 household projections.
Individual weights were post-stratified by region, locality, sex and age-groups(18-49, 50-59, 60-69, 70+) according to the 2009 Projected population estimates provided by Stats Ghana.
Weights are not normalised.