The 1997 Core Welfare Indicators Questionnaire (CWIQ) Survey is designed to furnish policy makers, planners and programme managers with a set of simple indicators for monitoring poverty and the effects of development policies, programmes and projects on living standards in the country. The survey also aims at providing reliable data on a timely basis for monitoring changes in the welfare status in various sub-groups of the population. An important feature of the survey is the use of Optical Mark Recognition (OMR) technology and high-speed scanners to generate statistical data rapidly.
The survey, which was carried out by the Ghana Statistical Service in collaboration with the World Bank marks Ghana's first experience in the application of data scanning technology to nation-wide surveys. Indeed, this is the first time such a survey has been successfully conducted in Africa.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
Household, Individual, Children under 5
The scope of Core Welfare Indicator Questionnaire Monitoring Survey 1997 includes:
Information on household members, Education, Health, Employment, Household assets, Househld amenities, Child roster (Children under 5).
health care and medical treatment [8.5]
general health [8.4]
Producers and sponsors
Ghana Statistical Service (GSS)
SAMPLE DESIGN FOR THE 1997 CWIQ MONITORING SURVEY
Objectives of the Sample Design
The primary objective of the sample design was to provide estimates with acceptable precision for monitoring poverty and the effects of development policies, programmes and projects on living standards in the country. The CWIQ sample also aimed at providing data on timely basis for monitoring changes in the welfare status in various sub-groups of the population.
The population was surveyed by designing a sample of households and collecting information on all members of the household.
The Ghana Statistical Service (GSS) maintains a complete list of Censal Enumeration Areas (EAs) with population and household information derived from the 1984 Population Census. This list, comprising a total of 12,969 EAs together with their respective household sizes constituted the sampling frame for the survey.
In order to improve the efficiency of the sample design, the sampling frame was classified into homogeneous strata. Specifically, tabulation of the survey results was done not only at the national level but also disaggregated by locality of residence (rural and urban). Each of the ten administrative regions also constituted a separate domain of estimation. Within these levels of stratification, the survey data was further disaggregated by poverty quintiles and socio-economic group of the head of household.
It was noted that monitoring implies comparing. Monitoring in time implies comparing the data from one year with the next. For monitoring in space, we examine the difference between areas in the same year. The crucial feature is that when two samples are compared, each one has an error variance and these two variances need to be added together.
Roughly, this implies that for monitoring changes in welfare status in various sub-groups of the population, we need samples about twice as large as when we are simply reporting an isolated result.
Now the most recent GLSS (Third Round) was based on a sample of 4,500 households. Analysis for this survey was at the national and regional levels. Following the above reasoning, a total sample size of 10,000 households was considered adequate for indicators at the national and regional levels. While it is important to ensure that the sample size is manageable operationally so as to control the quality of all the survey activities, it was considered that increasing these figures by roughly 50 percent would be a reasonable strategy and not excessively expensive using the CWIQ technology to improve the precision of the results. This would result in a total sample size of about 15,000 households.
Using the above sampling frame, the number of households to be selected per EA was based on the following reasoning. The Third Round of the Ghana Living Standards Survey used 10 households per EA in the rural sector and 15 in the urban sector. During the CWIQ Pilot Survey, 30 households were arbitrarily selected per EA to test the survey instruments. The overriding factor in the CWIQ is cost-effectiveness and simplicity and this favours a large take. However, the selection of 30 households per EA was considered too high. A sample size of 25 households per EA appeared more likely to be optimal.
Following the above parameters, the CWIQ survey was based on a two-stage, stratified, nationally representative cluster sample design.
Specifically, at the first stage, 588 Enumeration Areas (EAs) were selected using systematic sampling with probabilities proportional to size (PPS-method).
The distribution of the selected EAs by region is a follows:
REGION NUMBER OF EAs
Greater Accra 85
Brong Ahafo 57
Upper East 12
Upper West 26
A household listing exercise was carried out in the selected EAs. At the second stage, a systematic sample of 25 households per EA was selected. This sample design yielded a total ample of 14,700 households nationwide.
Note: Detail mathematical sample selection procedure is provided as external resource.
Dates of Data Collection
Data Collection Mode
Data Collection Notes
Mobile teams were used to collect the data. The data collection process lasted for 3 months, from September to November 1997. In all there were 13 teams, each comprising 4 enumerators, 1 supervisor and a driver. A senior-level supervisory team oversaw the implementation of the survey to ensure that high quality data were collected from the field.
Ghana Statistical Service
The questionnaire consists of only 4 doublesided forms. The CWIQ was designed to collect minimum amount of information needed to identify and classify target groups and to provide basic welfare monitoring information. Pre-coded multiple-choice response questionnaire were used. The questions were designed to be quicker and easier to administer and to process. Here are modules of this instrument:
A - GENERAL INFORMATION
B - INFORMATION ON HOUSEHOLD MEMBERS
C - EDUCATION
D - HEALTH
F - HOUSEHOLD ASSETS
G - HOUSEHOLD AMENITIES
H - POVERTY PREDICTORS
I - CHILD ROSTER (Children under 5)
For the first time in the history of the GSS, the survey was conducted using Optical mark Recognition (OMR) “bubble” questionnaires. To enter the data, these questionnaires were read by high speed scanners. The data processing team was able to perform simple on-line edit corrections while scanning. The data was then gradually transferred into a customized Access application for further, more complex validations. The Access application produced clean, validated and documented data files. This process took place at the same time the field work was on-going and finished only a couple of days after the last enumerator returned from the field. The clean data generated by the Access cleaning application was then processed through SPSS statistical package to produce a standardoutput bulletin within 12 days of the end of field work.
Ghana Statistical Service
Ghana Statistical Services
Ministry of Finance and Economic Planning
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download of the data files (for datasets obtained on-line)
Disclaimer and copyrights
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 Document ID
World Bank, Development Economics Data Group
Generation of DDI documentation
Date of Metadata Production
DDI Document version
Revised version of the DDI document "DDI-GHA-CWIQ-1997-WB-01" (Revised version of the DDI documentation which was imported from NSDStat prepared by Nesstar for the Africa Household Survey Databank in 2004).