National Household Survey 2002-2003
Uganda, 2002 - 2003
Created on
Apr 25, 2019
Last modified
Apr 25, 2019
Page views
341
Identification
Survey ID Number
UGA_2002_UNHS_v01_M
Title
National Household Survey 2002-2003
Country
| Name | Country code |
|---|---|
| Uganda | UGA |
Study type
Socio-Economic/Monitoring Survey [hh/sems]
Series Information
The Uganda National Household Survey 2002/03 was the eighth in a series of household surveys that started in 1988.
Abstract
The main objective of the Uganda National Household Survey 2002/03 was to collect high quality and timely data on demographic and socio-economic characteristics of household population for monitoring development performance of the country.
Specifically, the survey aimed at:
(a) Providing information on the economic characteristics of the population and its economic activity status i.e. the employment, unemployment and underemployment.
(b) Generating data for calculating gross output, value added, and other economic indicators required for National Accounts purposes.
(c) Integrating household socio-economic and community level surveys in the overall survey programme so as to provide an integrated data set. This will provide an understanding of the mechanisms and effects of various government programmes and policy measures on a comparative basis over time;
(d) Meeting special data needs of users for the Ministries of Health; Education; Gender, Labour and Social Development and other collaborating Institutions, together with donors and the NGO community so as to monitor the progress of their activities and interventions
(e) Generating and building social and economic indicators for monitoring the progress made towards social and economic development goals of the country
The UNHS 2002/03 was conducted in all districts except Pader. Some parts of Kitgum and Gulu districts were also not covered due to insecurity.
The survey included the following modules:
· Socio-economic module
· Labour force module
· Informal sector
· Community survey
Specifically, the survey aimed at:
(a) Providing information on the economic characteristics of the population and its economic activity status i.e. the employment, unemployment and underemployment.
(b) Generating data for calculating gross output, value added, and other economic indicators required for National Accounts purposes.
(c) Integrating household socio-economic and community level surveys in the overall survey programme so as to provide an integrated data set. This will provide an understanding of the mechanisms and effects of various government programmes and policy measures on a comparative basis over time;
(d) Meeting special data needs of users for the Ministries of Health; Education; Gender, Labour and Social Development and other collaborating Institutions, together with donors and the NGO community so as to monitor the progress of their activities and interventions
(e) Generating and building social and economic indicators for monitoring the progress made towards social and economic development goals of the country
The UNHS 2002/03 was conducted in all districts except Pader. Some parts of Kitgum and Gulu districts were also not covered due to insecurity.
The survey included the following modules:
· Socio-economic module
· Labour force module
· Informal sector
· Community survey
Kind of Data
Sample survey data [ssd]
Unit of Analysis
- Individual
- Household
- Community
- Household
- Community
Scope
Notes
The Socio-Economic module collected information on; characteristics of household members, health seeking behaviour of household members, prevention, channels of communication and HIV/AIDS, education and literacy, housing and household conditions, household consumption and non consumption expenditure, household and enterprise assets and welfare indicators.
The Labour Force module: This was for determining the total work force and deriving related parameters; current and usual activity status of household members, number of hours worked, previous employment, unemployment details, and the extent of child labour.
The Informal Sector module : This collected information about household enterprises and rural-based small-scale establishments. These are businesses undertaken by households with or without a fixed location. In addition, inputs and outputs of these enterprises for the major items were also collected. The components of the informal sector survey included; livestock, poultry, bee-keeping, and fishing; forestry; mining, quarrying and manufacturing; hotels, lodges and eating places; trade and services. In addition, the household survey investigated household and non-household based enterprises/establishments and was limited to: 1) Household based Enterprises in both rural and urban areas identified at listing stage; 2) Non-household based Enterprises in the rural areas. These were identified at the listing stage with assistance of the LC 1 guide.
The Community module: This module gathered information about the community (LC1). The information collected related to: Community characteristics; Community history and major events including access to and availability of social services namely schools, clinics, outlets for agricultural and non agricultural produce; Land tenure; Whether the community received the 25 percent Graduated Tax; Community projects undertaken in the three years preceding the survey and; Characteristic of the Education and Health infrastructure used by the community
The Labour Force module: This was for determining the total work force and deriving related parameters; current and usual activity status of household members, number of hours worked, previous employment, unemployment details, and the extent of child labour.
The Informal Sector module : This collected information about household enterprises and rural-based small-scale establishments. These are businesses undertaken by households with or without a fixed location. In addition, inputs and outputs of these enterprises for the major items were also collected. The components of the informal sector survey included; livestock, poultry, bee-keeping, and fishing; forestry; mining, quarrying and manufacturing; hotels, lodges and eating places; trade and services. In addition, the household survey investigated household and non-household based enterprises/establishments and was limited to: 1) Household based Enterprises in both rural and urban areas identified at listing stage; 2) Non-household based Enterprises in the rural areas. These were identified at the listing stage with assistance of the LC 1 guide.
The Community module: This module gathered information about the community (LC1). The information collected related to: Community characteristics; Community history and major events including access to and availability of social services namely schools, clinics, outlets for agricultural and non agricultural produce; Land tenure; Whether the community received the 25 percent Graduated Tax; Community projects undertaken in the three years preceding the survey and; Characteristic of the Education and Health infrastructure used by the community
Topics
| Topic | Vocabulary | URI |
|---|---|---|
| ECONOMICS [1] | CESSDA | http://www.nesstar.org/rdf/common |
| LABOUR AND EMPLOYMENT [3] | CESSDA | http://www.nesstar.org/rdf/common |
| TRADE, INDUSTRY AND MARKETS [2] | CESSDA | http://www.nesstar.org/rdf/common |
Keywords
| Keyword | Vocabulary |
|---|---|
| ECONOMICS [1] | CESSDA |
| LABOUR AND EMPLOYMENT [3] | CESSDA |
| TRADE, INDUSTRY AND MARKETS [2] | CESSDA |
Coverage
Geographic Coverage
The Uganda National Household Survey 2002/03 was conducted in all districts except Pader. Some parts of Kitgum and Gulu districts were also not covered due to insecurity.
Universe
The survey covered all resident population.
Producers and sponsors
Primary investigators
| Name | Affiliation |
|---|---|
| Uganda Bureau of Statistics (UBOS) | Ministry of Finance, Planning and Economic Development |
Funding Agency/Sponsor
| Name | Abbreviation | Role |
|---|---|---|
| Government of Uganda | GoU | Funding of operational cost |
| World Bank | WB | Funding of operational cost |
Other Identifications/Acknowledgments
| Name | Affiliation | Role |
|---|---|---|
| Respondents | Provided the required data | |
| Local Community leaders | Mobilising the communities | |
| Economic Policy Research Centre | Makerere University | Research Support |
| Institute of Statistics and Applied Economics of Makerere University | Makerere University | |
| Ministry of Gender, Labour and Social Development | Government of Uganda | |
| Population Secretariat | Ministry of Finance, Planning and Economic Development | |
| Uganda Manufacturer’s Association |
Sampling
Sampling Procedure
The UNHS sample was drawn through a stratified two-stage sampling design. The Enumeration Area (EA) was used as the first stage sampling unit and the household as the second stage-sampling unit. The sampling frame used for selection of first stage units (fsus) was the list of EAs with the number of households based on the cartographic work of the 2002 Population and Housing Census. A total of 972 EAs (565 in rural and 407 in urban areas) were covered. In order to select the second stage units, which are the households, a listing exercise using listing schedules was done in all selected EAs.
The sample size was determined by taking into consideration several factors, the three most important being: the degree of precision (reliability) desired for the survey estimates, the cost and operational limitations, and the efficiency of the design. UNHS 2002/03 covered a sample of 9,711 households.
Note: Details of the sampling design are given in Appendix III of the socio-economic survey report.
The sample size was determined by taking into consideration several factors, the three most important being: the degree of precision (reliability) desired for the survey estimates, the cost and operational limitations, and the efficiency of the design. UNHS 2002/03 covered a sample of 9,711 households.
Note: Details of the sampling design are given in Appendix III of the socio-economic survey report.
Response Rate
The response rate for the Uganda National Household Survey 2002/2003 was approximately 97%. A total of 9711 households were interviewed out of the 10,000 households initially targeted. A total of 289 households could not be interviewed mainly due to insecurity.
Weighting
Details of sampling weight calculation available in Section 3.2 of the Uganda National Household Survey 2002/2003 - Data Documentation.
Data Collection
Dates of Data Collection
| Start | End |
|---|---|
| 2002-05-03 | 2003-04-04 |
Data Collection Mode
Face-to-face [f2f]
Supervision
The Survey staff comprised of a total of 15 field teams.Four teams were recruited for each region except for Northern region which had three. Each field team was composed of one supervisor, four enumerators, and one driver. There was no gender bias and these supervisors and interviewers could be either male or female. For instance, out of 75 enumerators and supervisors, 30% were female. Each field supervisor was responsible for one team of Interviewers. During each round of fieldwork, each team supervisor had to write a report detailing what transpired on the respective trip. It included technical issues concerning the questionnaires where he/she could not take outright decision, coverage in terms of the EAs allocated, and the administrative issues. During each round of fieldwork, each team supervisor had to write a report detailing what transpired on the respective trip. It included technical issues concerning the questionnaires where he/she could not take outright decision, coverage in terms of the EAs allocated, and the administrative issues.
On the other hand, the team supervisor was tasked to make appointments with Local Council 1 (LC1) officials for the enumeration of their village, equitably distribute work amongst team members, check through the sampling process done by the enumerator, and check through questionnaires administered by the enumerators. He/she was also supposed to conduct an interview using the community questionnaire with LC1 officials and opinion leaders as respondents, interview the institution heads i.e. schools and health units most used by the community. Where applicable, he/she would also at times assist the enumerators in administering the household questionnaires. Repeated visits were made to those households where no 'well informed' member to be interviewed was found. In addition to the field teams, the survey management team as well as Bureau Staff at the level of officers and above undertook field supervision.
On average, each team was assigned two days to complete an enumeration area. An allowance of one day was allocated to teams that had to travel very long distances from the UBoS head office. For details on the number of field teams dispatched and the number of field days allocated for each survey month refer to the UBoS, Administrative Report, 2003. In the central office at the UBoS headquarters there were editing clerks who performed the data scrutiny before the data were sent for data entry. The field staff who did not go for field work during the respective field trips acted as the edit clerks.
On the other hand, the team supervisor was tasked to make appointments with Local Council 1 (LC1) officials for the enumeration of their village, equitably distribute work amongst team members, check through the sampling process done by the enumerator, and check through questionnaires administered by the enumerators. He/she was also supposed to conduct an interview using the community questionnaire with LC1 officials and opinion leaders as respondents, interview the institution heads i.e. schools and health units most used by the community. Where applicable, he/she would also at times assist the enumerators in administering the household questionnaires. Repeated visits were made to those households where no 'well informed' member to be interviewed was found. In addition to the field teams, the survey management team as well as Bureau Staff at the level of officers and above undertook field supervision.
On average, each team was assigned two days to complete an enumeration area. An allowance of one day was allocated to teams that had to travel very long distances from the UBoS head office. For details on the number of field teams dispatched and the number of field days allocated for each survey month refer to the UBoS, Administrative Report, 2003. In the central office at the UBoS headquarters there were editing clerks who performed the data scrutiny before the data were sent for data entry. The field staff who did not go for field work during the respective field trips acted as the edit clerks.
Data Collection Notes
The Survey staff comprised of a total of 15 field teams. Fieldwork was undertaken with the use of centrally recruited field teams whereby work in the sampled areas was programmed from the headquarters. There are four Statistical Regions, and the teams were recruited based on the languages most prevalent in each region. Four teams were recruited for each region. The data collection exercise started in May 2002 through April 2003 with a break in September 2002 due to the Census exercise.
Data Collectors
| Name | Abbreviation | Affiliation |
|---|---|---|
| Uganda Bureau of Statistics | UBOS | Ministry of Finance, Planning and Economic Development |
Questionnaires
Questionnaires
Nine types of questionnaires were used during the survey namely; Household Listing questionnaire, the Socio-Economic questionnaire, the Labourforce questionnaire, the Community questionnaire, Forestry Enterprise questionnaire, Trade and Services Enterprise questionnaire, Manufacturing, Mining and Quarrying Enterprise questionnaire, Livestock Enterprise questionnaire and Hotel Enterprise questionnaire. The last five questionnaires were administered to small-scale establishments and household enterprises. These were developed in consultation with various stakeholders. The household listing questionnaire was used to list all houses and households in the selected Enumeration Areas (EAs). Finally, the community questionnaire was administered at community level (Local Council level I).
Data Processing
Data Editing
A manual system of editing questionnaires was set up and a set of scrutiny notes to guide in manual checking was developed. In addition, range and consistency checks were included in the data-entry program. More intensive and thorough checks were carried out using MS-ACCESS by the processing team. Besides the editing done before data entry, the validation checks inbuilt in the program and double data entry, additional in-depth data cleaning on sections relevant for basic poverty analysis was done. For instance, individual level files were linked together to ensure that the same individual code reported in different sections of the questionnaire and in other modules corresponded to the same individual. Data cleaning on the other sections was also done. Any inconsistencies, data entry errors etc found were corrected after checking the original questionnaires.
Other Processing
CSPro was used to develop the data entry applications; Consistency and range checks were implemented in the CSPro data processing applications. Visual Basic was used to develop data Management Applications. These could automatically convert CSPro flat data files into Ms-Access, which was the final data storage tool. The system developed for data management could coordinate the sampled EAs, interviewed households, and entered households. It could also coordinate entered, compared, cleaned and converted data files. This left the data with very minimal completeness, consistence and data value errors. Other measures were put in place to minimize errors at the data entry stage. Data entry was done twice, that is, main entry and verification; there was 100 percent verification. A questionnaire could be considered clean after both datasets were perfectly comparable. When an entrant encountered a field problem, a field officer would be called upon to rectify the error.
Data Appraisal
Estimates of Sampling Error
There are two types of errors possible in any estimate based on a sample survey – sampling and non-sampling errors.
Non-sampling errors can be attributed to many sources which include: definitional difficulties, differences in the interpretation of questions by the interviewers, inability or unwillingness to provide correct responses on part of the respondents, mistakes in coding or recording the data, et cetera. Nonsampling errors would also occur in a complete census.
On the other hand, sampling errors occur because observations are made only on a sample, and not the entire population. Thus the accuracy of survey results is determined by the joint effects of the sampling and nonsampling errors.
For a given indicator, the sampling error is usually measured by the standard error. The standard error of a survey estimate is a measure of the variation among the estimates from all possible samples, and is a measure of the precision with which an estimate from a particular sample approximates the results from all possible samples. The accuracy of a survey result de pends on both the sampling and nonsampling error measured by the standard error and the bias; and other types of nonsampling errors not measured by the standard error.
The standard errors of the rates presented in this appendix were computed using the SASÓ PROC SURVEYMEANS procedure. This procedure does not assume that the data was taken from a simple random sample, but rather from a more complex design. The SurveyMeans Procedure takes into account the effect of clustering and stratifying in the calculation of the variances and standard errors, using the Taylor expansion method to estimate these sampling errors.
The sampling errors are computed for selected variables considered to be of interest, but can be computed for all variables in the dataset. The sampling errors are presented for the country as a whole, for women and men where relevant, and for rural and urban areas and for each of the four regions: Central, East, West and North. For each variable the type of statistic (mean, sum, rate) are given as well as the standard error, the 95% confidence limits, and the coefficient of variation.
Generally the standard errors of most national estimates are small and within acceptable limits, but there is wider variability for the estimates of the subpopulations. For example for the Net Attendance Ration (NER), the standard error for the whole country is 6.5 percent, while for urban and rural areas it is 7.6 and 7.3 percent respectively. For more details about the estimates of sampling error can be found in Appendix IV of "UNHS 2002/2003 Report on the Socio-Economic Survey"
Non-sampling errors can be attributed to many sources which include: definitional difficulties, differences in the interpretation of questions by the interviewers, inability or unwillingness to provide correct responses on part of the respondents, mistakes in coding or recording the data, et cetera. Nonsampling errors would also occur in a complete census.
On the other hand, sampling errors occur because observations are made only on a sample, and not the entire population. Thus the accuracy of survey results is determined by the joint effects of the sampling and nonsampling errors.
For a given indicator, the sampling error is usually measured by the standard error. The standard error of a survey estimate is a measure of the variation among the estimates from all possible samples, and is a measure of the precision with which an estimate from a particular sample approximates the results from all possible samples. The accuracy of a survey result de pends on both the sampling and nonsampling error measured by the standard error and the bias; and other types of nonsampling errors not measured by the standard error.
The standard errors of the rates presented in this appendix were computed using the SASÓ PROC SURVEYMEANS procedure. This procedure does not assume that the data was taken from a simple random sample, but rather from a more complex design. The SurveyMeans Procedure takes into account the effect of clustering and stratifying in the calculation of the variances and standard errors, using the Taylor expansion method to estimate these sampling errors.
The sampling errors are computed for selected variables considered to be of interest, but can be computed for all variables in the dataset. The sampling errors are presented for the country as a whole, for women and men where relevant, and for rural and urban areas and for each of the four regions: Central, East, West and North. For each variable the type of statistic (mean, sum, rate) are given as well as the standard error, the 95% confidence limits, and the coefficient of variation.
Generally the standard errors of most national estimates are small and within acceptable limits, but there is wider variability for the estimates of the subpopulations. For example for the Net Attendance Ration (NER), the standard error for the whole country is 6.5 percent, while for urban and rural areas it is 7.6 and 7.3 percent respectively. For more details about the estimates of sampling error can be found in Appendix IV of "UNHS 2002/2003 Report on the Socio-Economic Survey"
Access policy
Access authority
| Name | Affiliation | URL | |
|---|---|---|---|
| Uganda Bureau of Statistics (UBOS) | Ministry of Finance, Planning and Economic Development | ubos@ubos.org | www.ubos.org |
Contacts
| Name | Affiliation | URL | |
|---|---|---|---|
| Uganda Bureau of Statistics (UBOS) | Ministry of Finance, Planning and Economic Development | ubos@ubos.org | www.ubos.org |
Confidentiality
Confidentiality of respondents is guaranteed by Article 19 of The Uganda Bureau of Statistics Act, 1998. Before being granted access to the dataset, all users have to formally agree:
1. To make no copies of any files or portions of files to which s/he is granted access except those authorized by the data depositor.
2. Not to use any technique in an attempt to learn the identity of any person, establishment, or sampling unit not identified on public use data files.
3. To hold in strictest confidence the identification of any establishment or individual that may be inadvertently revealed in any documents or discussion, or analysis. Such inadvertent identification revealed in her/his analysis will be immediately brought to the attention of the data depositor.
Access conditions
Metadata is provided to you for exclusive use. The metadata may not be transferred to any other user without prior written permission from UBOS.
Citation requirements
The following statement must be used as citation " Source of data: Uganda Bureau of Statistics, Uganda National Household Survey 2002/2003, Version 1.0 of the data set (November 2006), provided by UBOS."
Disclaimer and copyrights
Disclaimer
The Uganda Bureau of Statistics (UBOS) provides these metadata to external users without any warranty or responsiblity implied. UBOS accepts no responsibility for the results and/or implications of any actoins resulting from the use of these data.
Copyright
(c) 2008, Uganda Bureau of Statistics
Metadata production
DDI Document ID
DDI_UGA_2002_UNHS_v01_M
Producers
| Name | Abbreviation | Affiliation | Role |
|---|---|---|---|
| Uganda Bureau of Statistics | UBOS | Ministry of Finance, Planning & Economic Development | Documentation of the survey |
Date of Metadata Production
2009-11-30
DDI Document version
Version 01: Adopted from "DDI-UBOS-UNHS2002-2003/02 - v1.0" DDI that was done by metadata producers mentioned in "Metadata Production" section.