Ghana - Ghana Living Standards Survey 3 -1991, Third round
Reference ID | GHA-GSS-GLSS 3-1991-v2.1 |
Year | 1991 - 1992 |
Country | Ghana |
Producer(s) | Ghana Statistical Service (GSS) - Office of the President |
Sponsor(s) | Goverment of Ghana - GoG - Logistical assistance The World Bank - - Support in all aspect of the survey European Union - EU - Support in all aspect of the survey |
Metadata | Documentation in PDF |
Created on
Feb 20, 2009
Last modified
Mar 12, 2016
Page views
1922416
Sampling
Sampling Procedure
A multi-stage sampling technique was used in selecting the GLSS3 sample. Initially, 4565 households were selected for GLSS3.This spread around the country in 407 enumeration areas (EAs). 15 households were taken in an urban EA and 10 households in a rural EA. The actual achieved sample was 4552 households. The sample designed used for the GLSS3, and with the very high response rate achieved, the sample can be considered as being self-weighting, though in the case of expenditure data weighting of the expenditure values should be done.
Deviations from Sample Design
There was no deviation from the sampling design.
Response Rate
Out of the selected 4565 households 4552 were successfully interviewed . Details can be found in Appendix 1. of the main report.
Weighting
The design chosen for this survey was intended to ensure that the sample was self-weighting, provided an adequate response was secured on the survey. The number of households achieved in the survey almost exactly matched the number required. The small shortfall arose in cases where an interviewer failed to secure interviews from several of the selected households, and then used up all the reserve list of households in the EA without managing to complete the quota. The number of such cases is so few that it has not been considered necessary to do any imputations or re-weighting for missing households. However there was re-weighting of income and expenditure. The reason for this is given in appendix 1. of main report.