The Uganda Demographic and Health Survey (UDHS) was conducted by the Ministry of Health in 24 districts between September 1988 and February 1989. The sample covered 4730 women aged 15-49. Nine northern districts were not surveyed due to security reasons. The purpose of the survey was to provide planners and policymakers with baseline information regarding fertility, family planning, and maternal and child health. The survey data were also needed by UNFPA and UNICEF- Kampala for planning and evaluation of current projects in Uganda.
The primary objective of the UDHS was to provide data on fertility, family planning, childhood mortality and basic indicators of maternal and child health. Additional information was collected on educational level, literacy, sources of household water and housing conditions. The available demographic data were incomplete and hardly any recent information concerning family planning or other health and social indicators existed at the national level.
A more specific objective was to provide baseline data for the South West region and the area in Central region known as the Luwero Triangle, where the Uganda government and UNICEF are currently supporting a primary health care project. In order to effectively plan strategies and to evaluate progress in meeting the project goals and objectives, there was a need to collect data on the health of the target population.
Another important goal of UDHS was to enhance the skills of those participating in the project so that they could conduct high-quality surveys in the future. Finally, the contribution of Ugandan data to an expanding international data set was an objective of the UDHS.
Kind of Data
Sample survey data
Unit of Analysis
- Women age 15-49
The Uganda Demographic and Health Survey (UDHS) was conductedin 24 districts. Nine northern districts were not surveyed due to security reasons.
Unit of Analysis
- Women age 15-49
The population covered by the 1988 UDHS is defined as the universe of all women age 15-49 in Uganda and all men age 15-54 living in the household. But due to security problems at the time of sample selection, 9 districts, containing an estimated 20 percent of the country's population, were excluded from the sample frame
Producers and sponsors
Authoring entity/Primary investigators
Ministry of Health
Ministry of Planning and Economic Development
Department of Geography, Makerere University
Institute of Statistics and Applied Economics
Institute for Resource Development/Macro Systems
U.S. Agency for International Development
The UDHS used a stratified, weighted probability sample of women aged 15-49 selected from 206 clusters. Due to security problems at the time of sample selection, 9 districts, containing an estimated 20 percent of the country's population, were excluded from the sample frame. Primary sampling units in rural areas were sub-parishes, which, in the absence of a more reliable sampling frame, were selected with a probability proportional to the number of registered taxpayers in the sub-parish. Teams visited each selected sub-parish and listed all the households by name of the household head. Individual households were then selected for interview from this list.
Because Ugandans often pay taxes in rural areas or in their place of work instead of their place of residence, it was not possible to use taxpayer rolls as a sampling frame in urban areas. Consequently, a complete list of all administrative urban areas known as Resistance Council Ones (RCls) was compiled, and a sampling frame was created by systematically selecting 200 of these units with equal probability. The households in these RCls were listed, and 50 RCls were selected with probability proportional to size. Finally, 20 households were then systematically selected in each of the 50 RCls for a total of 1,000 urban households.
The sample used for the Uganda Demographic and Health Survey was a stratified, weighted probability sample of women aged 15-49 selected from 206 clusters. Due to security problems at the time of sample selection, 9 of the country's 34 districts, containing an estimated 20 percent of the population, were excluded from the sample frame. Primary sampling units in rural areas were sub-parishes, which, in the absence of a more reliable sampling frame, were selected with a probability proportional to the number of registered taxpayers in the sup-parish.
The South West region and the area in Central region known as Luwero Triangle were each over-sampled to provide a sample with sufficient size to produce independent estimates of certain variables for these two areas.
The urban sector was over-sampled by a factor of three compared with a proportionate urban/rural sample. Since it was not possible to use an appropriate sampling frame in the urban area, it was necessary to look for an altemative procedure. A convenient solution avoiding excessive cost was to use a two-phase sampling:
- 1st Phase: A complete list of all administrative urban areas known as Resistance Council Ones (RCls) was compiled and a sampling frame was created by systematically selecting 200 of these units with equal probability for a complete household updating.
- 2nd Phase: After the first phase selection and updating was completed, a sub-sample of 50 RCls were selected with probability proportional to size (size as reported in the housing listing). At the subsequent stage, 20 households were then systematically selected in each of the 50 RCls for a total of 1,000 urban households.
Deviations from the Sample Design
Contact was not made with 127 eligible women, either because the respondent was not at home during any of the visits by the interviewer, or because the respondent refused to be interviewed, or because of other reasons. In any case, the overall level of nonresponse is very low.
Households and eligible women: Out of 5,587 addresses visited, 5,123 households were located. The remaining addresses (8.3 percent) were not valid households, either because the dwelling had been vacated or destroyed, or the household could not be located or did not exist. Of the located households, 5101 were successfully interviewed, producing a household response rate of 99.6 percent.
The household questionnaires identified 4,857 women eligible for the individual interview (that is, they were aged 15-49 and had spent the night before the interview in the selected household). This represents an average of slightly under one eligible women per household. Questionnaires were completed for 4,730 women, indicating an individual response rate of 98.4 percent. The overall response rate, that is, the product of response rates at the household and individual levels was 98.0 percent
The response rates for the urban-rural areas, and regions were similar. In the urban areas, the overall individual response rate was 96.0 percent, compared with 97.7 percent for the rural areas. These lower rates of response in the urban areas are influenced by the low rates of response observed for Kampala.
Dates of Data Collection (YYYY/MM/DD)
Mode of data collection
Type of Research Instrument
Three questionnaires were used for the UDHS: the household questionnaire, the individual woman's questionnaire, and the service availability questionnaire.
a) The household questionnaire listed all usual members of the household and their visitors, together with information on their age and sex and information on the fostering of children under 15. It was used to identify women who were eligible for the individual interview, namely, those aged 15-49 who slept in the household the night before the household interview, whether they normally lived there or were visiting.
b) For those women who were either absent or could not be interviewed during the first visit, a minimum of three revisits were made before recording nonr esponse. Women were interviewed with the individual questionnaire, which contained questions on fertility, family planning and maternal and child health.
c) The service availability (SA) questionnaire collected information on family planning and health services and other socioeconomic characteristics of the selected areas and was completed for each rural cluster and for each urban area. The SA questionnaire was administered by a different team of interviewers from the one carrying out the individual women's interview. The same clusters chosen for the individual interviews were visited by the SA interviewer who was instructed to assemble 3 or 4 "knowledgeable" residents. These people were asked about the services available in the community and the distances to them. Based on this information, interviewers visited the facilities close to the cluster and collected information about equipment, staffing, services available, and general infrastructure. Results on service availability are not included in this report.
The household and the individual questionnaires were translated into four languages: Luganda, Lugbara, Runyankole-Rukiga and Runyoro-Rutom. Luganda questionnaires were used in the East region, where there are a number of languages, but most people speak Luganda. A pretest of the translated questionnaires was conducted in October 1987 by interviewers who completed a three-week training course.
Ministry of Health
Completed questionnaires were sent to the data processing room at Makerere University where data entry and machine editing proceeded concurrently with fieldwork. Four desktop computers and ISSA, the Integrated System for Survey Analysis, were used to process the UDHS data. Of the households sampled, 5,101 were successfully interviewed, a completion rate of 91.3 percent. A total of 4,857 eligible women were identified in these households, of which 4,730 were interviewed, a completion rate of 97.4 percent. Data entry and editing were completed a few days after fieldwork ended.
Nonsampling error is due to mistakes made in carrying out field activities, such as failure to locate and interview the correct household, errors in the way questions are asked, misunderstanding of the questions on the part of either the interviewer or the respondent, data entry errors, etc. Although efforts were made during the design and implementation of the UDHS to minimize this type of error, nonsampling errors are impossible to avoid and difficult to evaluate analytically.
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
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.