Identification
Survey ID Number
LSO_2002_CWIQ_v01_M
Title
Core Welfare Indicators Questionnaire 2002
Country
| Name | Country code |
|---|---|
| Lesotho | LSO |
Study type
Core Welfare Indicators Questionnaire [hh/cwiq]
Abstract
The Core Welfare Indicator Questionnaire (CWIQ) Survey is part of the efforts being made by the Bureau of Statistics (BOS) to provide Statistical Information timely to the government as part of the framework for monitoring Poverty Reduction programmes. The CWIQ technique had been developed by the World Bank in collaboration with other development partners and it was to circumvent the problem of delayed processing of African Surveys. CWIQ uses the technology of optical scanner to capture the survey data and thereby fastens the data processing and hence quick publication of the results.
The objectives of the current National CWIQ Survey were:
(i) Elaborating main indicators for social welfare and basic needs of the socio-economic groups of the population.
(ii) Identifying target groups for benefiting special action programs designed by decision makers to address their needs.
(iii) Monitoring changes happening in the welfare of the households overtime.
(iv) Providing a database for social research.
(v) Elaborating on numerous sector programs aimed at improving the welfare of the population across the country. In order to prepare these programs, it is always necessary to identify the problems to be addressed by the policies and to know to which extent the population is affected by these problems.
(vi) Building up models to stimulate the global impact of policy choices and the distribution of the impact.
The objectives of the current National CWIQ Survey were:
(i) Elaborating main indicators for social welfare and basic needs of the socio-economic groups of the population.
(ii) Identifying target groups for benefiting special action programs designed by decision makers to address their needs.
(iii) Monitoring changes happening in the welfare of the households overtime.
(iv) Providing a database for social research.
(v) Elaborating on numerous sector programs aimed at improving the welfare of the population across the country. In order to prepare these programs, it is always necessary to identify the problems to be addressed by the policies and to know to which extent the population is affected by these problems.
(vi) Building up models to stimulate the global impact of policy choices and the distribution of the impact.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
- Households
- Individuals
- Children
- Individuals
- Children
Scope
Notes
The 2002 Lesotho Core Welfare Indicators Questionnaire covered following topics:
- Demographic characteristics of household members
- Education of household members
- Health of household members
- Employment of household members
- Household assets
- Household amenities
- Poverty predictors
- Child nutrition through anthropometric measurements
- HIV/AIDS module
- Demographic characteristics of household members
- Education of household members
- Health of household members
- Employment of household members
- Household assets
- Household amenities
- Poverty predictors
- Child nutrition through anthropometric measurements
- HIV/AIDS module
Coverage
Geographic Coverage
The survey covered all de jure household members (usual residents) through out the country.
Universe
National
Producers and sponsors
Primary investigators
| Name | Affiliation |
|---|---|
| Bureau of Statistics (BoS) | Ministry of Finance and Development Planning |
Producers
| Name | Role |
|---|---|
| The World Bank | Technical assistance |
Funding Agency/Sponsor
| Name | Abbreviation | Role |
|---|---|---|
| Government of Lesotho | ||
| The World Bank | WB | |
| East African Statistical Training Centre | EASTC | Buying of scanners for data processing |
Sampling
Sampling Procedure
The CWIQ main survey was conducted as a module of national household survey capability programme (NHSCP) run by the Bureau of Statistics. The desire had been to keep the design as simple as possible, so a two-stage sample design was employed. Enumeration Areas (E.A.s) were the first stage sampling units and the Households (HH's) were the second stage sampling units. A minimum of 400-500 households represented an adequate sample size to give reliable domain estimates. Therefore for each domain 25 clusters were advocated with a “take” of 20 households per cluster. However for Maseru urban an additional 200 households in 10 clusters were taken. The total national sample was made up of:
a. (500*10) = 5000 households.
b. Maseru urban additional 200 households in 10 clusters. Total was 5200 households.
In each district the sample was distributed over rural and urban areas in the ratio of 2:1 respectively so as to obtain reasonable sample sizes in the urban to give urban/rural estimates in each district. Currently the rural/urban distribution is about 4:1.
The design was to select the E.A.s rather than the PSUs, which were selected during the pilot survey. The reason was that the PSUs were too large, thus making supervision very difficult. Additional costs were incurred in traveling within the PSU in terms of time and fuel consumption. This resulted in the extended time to finish the fieldwork. The E.A.s were more compact clusters and most of the problems experienced in the pilot were eliminated.
Since the E.A.s were delineated for the 1996 population census, household listing was absolutely essential before the selection of the households. It was more convenient and cost effective to list the households within an E.A. rather than listing in the whole Psu as was the case in the pilot survey.
The master sample was a proportional allocation of the PSU's to the districts. The design according to the master sample would have made the design unsuitable for district estimates as some districts notably Quthing, Qacha's Nek, Mokhotlong and Thaba-Tseka would have inadequate sample sizes to give reliable estimates at the district level.
In order to have the same margin of error for all districts, the design had aimed at selecting equal number of E.A.s for each district.
a. (500*10) = 5000 households.
b. Maseru urban additional 200 households in 10 clusters. Total was 5200 households.
In each district the sample was distributed over rural and urban areas in the ratio of 2:1 respectively so as to obtain reasonable sample sizes in the urban to give urban/rural estimates in each district. Currently the rural/urban distribution is about 4:1.
The design was to select the E.A.s rather than the PSUs, which were selected during the pilot survey. The reason was that the PSUs were too large, thus making supervision very difficult. Additional costs were incurred in traveling within the PSU in terms of time and fuel consumption. This resulted in the extended time to finish the fieldwork. The E.A.s were more compact clusters and most of the problems experienced in the pilot were eliminated.
Since the E.A.s were delineated for the 1996 population census, household listing was absolutely essential before the selection of the households. It was more convenient and cost effective to list the households within an E.A. rather than listing in the whole Psu as was the case in the pilot survey.
The master sample was a proportional allocation of the PSU's to the districts. The design according to the master sample would have made the design unsuitable for district estimates as some districts notably Quthing, Qacha's Nek, Mokhotlong and Thaba-Tseka would have inadequate sample sizes to give reliable estimates at the district level.
In order to have the same margin of error for all districts, the design had aimed at selecting equal number of E.A.s for each district.
Response Rate
Out of a total sample of 5200 households, 4954 respondents responded giving a response rate of about 95.3. The 4.7% non-response was made up of “not found” households and a few refusals. The rural response rate was 97.8% as against 94.7% in the urban.
Weighting
No weights provided for the data set.
Data Collection
Dates of Data Collection
| Start | End |
|---|---|
| 2002-04-22 | 2002-05-23 |
Data Collection Mode
Face-to-face [f2f]
Supervision
Apart from the training, two other measures were set in place to control the quality of data, especially at the data collection stage. The first concerned the work of the supervisors apart from coordinating the work of their individual groups; they also had the responsibility for quality control. This they did through direct observation of and assistance given to the enumerators under them and also by editing or reviewing completed questionnaires before finally submitting them for data processing.
The other measure was the setting up of a monitoring team made up of two coordinators and senior survey managers from the head office. Armed with guidelines members of the team were in the field throughout the fieldwork period, monitoring the data collection activities. The monitoring team met once a week and when necessary (during the fieldwork period) to review their findings and map out strategies to deal with any lapses observed. The monitoring exercise no doubt contributed to the overall quality of the fieldwork.
The other measure was the setting up of a monitoring team made up of two coordinators and senior survey managers from the head office. Armed with guidelines members of the team were in the field throughout the fieldwork period, monitoring the data collection activities. The monitoring team met once a week and when necessary (during the fieldwork period) to review their findings and map out strategies to deal with any lapses observed. The monitoring exercise no doubt contributed to the overall quality of the fieldwork.
Data Collection Notes
Training
An 11-day training workshop was conducted for the staff (field staff and data processing staff). The training was quite detailed and intensive. The trainees were taken through the general interview process of the CWIQ, interviewing techniques, how to complete the questionnaire, sample selection, mock interviews and field practices. There was also a training and demonstration on how to carry out anthropometric measurements. The mock interview and field practices provided the field-staff the opportunities for practicing how to “bubble” the CWIQ. Series of tests were also conducted for the trainees. The tests merely strengthened the understanding of the trainees but were not used to select staff, as there were not enough to choose from. The supervisors were separately trained for 3 days briefing them on their duties, which included questionnaire editing and supervision. The training was a residential one held in Roma.
An 11-day training workshop was conducted for the staff (field staff and data processing staff). The training was quite detailed and intensive. The trainees were taken through the general interview process of the CWIQ, interviewing techniques, how to complete the questionnaire, sample selection, mock interviews and field practices. There was also a training and demonstration on how to carry out anthropometric measurements. The mock interview and field practices provided the field-staff the opportunities for practicing how to “bubble” the CWIQ. Series of tests were also conducted for the trainees. The tests merely strengthened the understanding of the trainees but were not used to select staff, as there were not enough to choose from. The supervisors were separately trained for 3 days briefing them on their duties, which included questionnaire editing and supervision. The training was a residential one held in Roma.
Questionnaires
Questionnaires
The main survey instruments used for data collection were the generic CWIQ designed by the World Bank with some few modifications. A survey manual appropriately modified (also developed by the World Bank) served as a complement to the CWIQ and was in fact the main instructional manual used in training the staff. The measuring boards and measuring scales were used to measure the height and weight of the children respectively. Notebooks were supplied to the field staff to record the experiences that affected their work.
Data Processing
Other Processing
The CWIQ questionnaire was designed for data to be extracted from it using optical scanning. The responses were recorded in the questionnaire by filling in bubbles assigned to each question. During scanning, the scanner created an image of each page of the questionnaire that is then evaluated by the scanning software. A scanning operator who verified entries and corrects errors in the extracted data further verified the input. The data is then converted into a database format where data validation, correction and tabulation were done.
The data processing system for the Lesotho CWIQ national survey was adapted from the generic CWIQ system to incorporate some changes specific to Lesotho in the questionnaire. The data processing office was set up in the Demographic, Labour and Social Statistics division of the Bureau of Statistics (BOS) in Maseru. Office staff consisted of two programmers and five data entry staff under the direction of a consultant. Three laptops, a desktop computer, an optical scanner and a laser printer were employed in the scanning and verification of questionnaires, in the correction, validation and tabulation of the data.
On arrival in the data processing office, the cluster and household numbers of each questionnaire was confirmed using a master list. There was continuous monitoring of the logs throughout the processing to compare them with the computer tally because the questionnaires were not arriving as complete clusters and had to be processed in parts.
The received questionnaires were counted and entered in the data processing log. They were then prepared for scanning by the data entry operators by counting, sorting them and removing staples. Checking for completeness of the first page and the reference numbers were on all pages was also done.
After scanning and verification, the data was transferred to the database where validation programs were run and errors printed. Corrections were made on the validation error printouts and these were used to correct data in the database. The running of the validation programs, printing and correction of errors was repeated until all the errors were removed, and until all the questionnaires were returned from the field. The process finished one and half weeks after the last questionnaires were returned. A last validation was done on all the data ensuring that only exceptions agreed to were left in the error output.
The database was updated with a set of analysis variables derived from data in the questionnaires such as type of residence (urban/rural), characteristics of the household head, household size, welfare quintiles, nutrition indicators for children under five and the household weight to be used for aggregate result.
Data summaries and standard tables were produced after 2 days. Modifications to some tables were also done as well as figures of some of the tables. Lastly, sampling errors of the core welfare indicators were produced.
The data processing system for the Lesotho CWIQ national survey was adapted from the generic CWIQ system to incorporate some changes specific to Lesotho in the questionnaire. The data processing office was set up in the Demographic, Labour and Social Statistics division of the Bureau of Statistics (BOS) in Maseru. Office staff consisted of two programmers and five data entry staff under the direction of a consultant. Three laptops, a desktop computer, an optical scanner and a laser printer were employed in the scanning and verification of questionnaires, in the correction, validation and tabulation of the data.
On arrival in the data processing office, the cluster and household numbers of each questionnaire was confirmed using a master list. There was continuous monitoring of the logs throughout the processing to compare them with the computer tally because the questionnaires were not arriving as complete clusters and had to be processed in parts.
The received questionnaires were counted and entered in the data processing log. They were then prepared for scanning by the data entry operators by counting, sorting them and removing staples. Checking for completeness of the first page and the reference numbers were on all pages was also done.
After scanning and verification, the data was transferred to the database where validation programs were run and errors printed. Corrections were made on the validation error printouts and these were used to correct data in the database. The running of the validation programs, printing and correction of errors was repeated until all the errors were removed, and until all the questionnaires were returned from the field. The process finished one and half weeks after the last questionnaires were returned. A last validation was done on all the data ensuring that only exceptions agreed to were left in the error output.
The database was updated with a set of analysis variables derived from data in the questionnaires such as type of residence (urban/rural), characteristics of the household head, household size, welfare quintiles, nutrition indicators for children under five and the household weight to be used for aggregate result.
Data summaries and standard tables were produced after 2 days. Modifications to some tables were also done as well as figures of some of the tables. Lastly, sampling errors of the core welfare indicators were produced.
Access policy
Contacts
| Name | Affiliation | |
|---|---|---|
| Head, Division of Demography, Labour and Social Statistics | Bureau of Statistics | demography@bos.gov.ls |
Confidentiality
Confidentiality of respondents is guaranteed by Statistics Act of 2001.
Citation requirements
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
- 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
Disclaimer and copyrights
Disclaimer
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.
Metadata production
DDI Document ID
DDI_LSO_2002_CWIQ_v01_M
Producers
| Name | Abbreviation | Role |
|---|---|---|
| World Bank, Development Economics Data Group | DECDG | Production of metadata |
DDI Document version
Version 01: (July 2011)