GMB_1992_HES_v01_M
Household Economic Survey 1992
Name | Country code |
---|---|
Gambia, The | GMB |
Income/Expenditure/Household Survey [hh/ies]
The Household Economic Survey is an important component of the Social Di mensions of Adjustment (SDA) program. The Household Economic Survey is a large and complex instrument. There are 14 sections in all, dealing with a range of household and individual information. Despite its size it's only a subset of the full integrated survey proposed by the World Bank.
The Household Economic Survey is an important component of the Social Dimensions of Adjustment (SDA) program. It is designed to provide social and economic data on the welfare of households following the Introduction of World Bank programs for economic reconstruction. The focus of the survey is therefore diagnostic -explaining how and why households respond to changes in the mesoeconomic environment and how their well-being is thereby affected. One of the key objectives of the household survey is to provide indicators for different socio-economic categories of household defined as Socio Economic Groupings [SEGs], particularly the poorer households.
Sample survey data [ssd]
Households and Individuals
The 1992 Gambia Household Economic Survey covered the following topics:
Household particulars
Household roster
Employment
Migration
Crop production
Livestock
Non-farm enterprise
Housing
Consumption of own produce
Household expenditure -- food items
Miscellaneous income and expenditure
Transfer payments made by household
Transfer payments received by household
Anthropometry
National
Name | Affiliation |
---|---|
Central Statistics Department | Ministry of Finance and Economic Affairs |
Name | Role |
---|---|
The World Bank | Technical assistance |
Name | Role |
---|---|
The World Bank | Funding |
United Nations Development Programme | Funding |
To meet the objectives outlined in the Introduction the survey needed to cover a sufficiently large number of house holds selected in a statistically reliable manner. Overall sampling and budgetary considerations suggested that a sample size of about 1400 households would be both statistically appropriate and financially feasible. It would be statistically appropriate because it would provide more than enough cases for a national sample and sufficient cases for Divisional level analysis. It was appropriate to the budget because estimates of the time and resources suggested it was well within the capabilities of the team envisaged for data collection.
It is technically possible to draw a simple random sample from all of the 82,000 households in Gambia. However it is not economically feasible to conduct such a survey because of the large amount of travel that would be required to conduct the interviews in rural areas with a scattered population. Therefore some method of clustering the households was necessity to provide for a staged sampling procedure. Geographical clustering already exists in the form of census Enumeration Areas [EAs]. These EAs are mapped to contain approximately 500 persons, and cover the entire county, conforming to the administrative boundaries. Enumeration Areas are of approximately the same size (500 persons). However in actuality they range from about 300 to 1000 persons. Some classification by size is desirable to maintain sampling probabilities.
The number of households selected per EA is a further factor in the sampling process. Maximizing the number of households per EA has the advantage of reducing travel costs. It also increases sampling error by sharply reducing the number of EAs sampled. Minimizing the number of households per EA greatly in creases costs but does not affect sampling error to the same extent. A constant take of households per EA has no effect on the sampling error over proportional probability sampling in stage one. Because urban populations are more likely to be residentially homogenous [poor people live In the same district; rich people similarly live in their own districts I the constant take for urban EAs is set at half of that for rural EAs. In villages the rich and the poor are more likely to be found within the same EA.
Taking all the above considerations into account it was decided to use a multi- stage sampling approach using probability proportional to size as recommended. The base cluster would be the Enumeration Areas defined in the 1983 Population Census. The stages would take into ac count administrative boundaries and population density.
The survey was collected using two questionnaires:
Start | End |
---|---|
1992-11 | 1993-03 |
Field Work - Training
All supervisors, interviewers and data entry clerks went through four weeks of training on data collection. The training included interview techniques, detailed discussion of each question, and training in measuring and estimating quantities consumed for the consumption of own produce section.
Because the majority of interviews would be conducted in one of the local language some time was spent on ensuring standard translations of the key questions. It was anticipated that most interviews would be conducted in Mandinka, Wollof or Fula the three most common local languages. Interviewers were instructed to secure an interpreter if there was any need.
The trainees conducted some household interviews under close supervision in the Greater Banjul area and also in the North Bank Division which is largely rural and agricultural. The data entry clerks collected data in Greater Banjul for a month, then they received further training in the specifics of the data entry program.
Data Collection
The data was collected from the beginning of November 1992 to the end of March 1993. In rural areas a field team conducted roughly a round of Interviews in two EAs (36 interviews) per week. As the team had to conduct two rounds of interviews two weeks apart this means that a team spent roughly one week alto gather in a rural EA. The field teams were based in five locations around the country. Households were defined as a group of persons acknowledging one head and with some sharing of food and budgets. In the Gambian context this meant that most polygamous households were counted as one large household.
The data entry took place in the head office in Banjul, where the process was supervised by senior staff. Data entry used the US Bureau of Census program IMPS, which provides extensive facilities for data entry and checking. The surveys were extensively preceded and the data entry operators referred any question able data back to one of the office super visors. One of the advantages of the IMPS system is its ability to produce concatenated hatches easily and to process frequency tables using the data dictionary defined for data entry. It was therefore possible to have frequent updates of the data entered and check for trends and obvious errors. The data entry operators were able to maintain a good speed of data entry.
Data Cleaning
Because of the precoded data entry program there were few out of range errors in the data. Most of the data cleaning process was involved with enuring that each household was represented in the seventeen datasets that comprised the complete run of data. Some households were duplicated and some had not been collected, or not returned after call backs.
There were some errors in mispunched legitimate codes but on the whole the rigorous program of checking at several stages before data entry kept the reliability and integrity of the data. Due to the mere size of the sections on expenditure and consumption (these sections contained about 220,000 records in total), this part of the cleaning was very time consuming.
The use of the datasets must be acknowledged using a citation which would include:
Example:
Gambia Central Statistics Department. Gambia Household Economic Survey (HES) 1992. Ref. GMB_1992_HES_v01_M. Dataset downloaded from [URL] on [date].
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.
DDI_GMB_1992_HES_v01_M_WB
Name | Affiliation | Role |
---|---|---|
Development Data Group | World Bank | Generation of DDI documentation |
2013-06-12
Version 1.0 (June 2013)