NGA_1999_DHS_v01_M
Demographic and Health Survey 1999
Name | Country code |
---|---|
Nigeria | NGA |
Demographic and Health Survey (standard) - DHS IV
The 1999 Demographic and Health Survey (NDHS) is the second survey of its kind in Nigeria. The first one was conducted in 1990.
Sample survey data
The Nigeria Demographic and Health Survey 1999 covers the following topics:
The 1999 Nigeria Demographic and Health Survey (NDHS) is a nationally representative survey. The sample was stratified into rural and urban areas and was selected in two stages. It was designed to produce reliable estimates of most of the variables for the rural and urban segments of the country as well as each of five statistical regions, namely, the Northeast region, the Northwest region, the Central region, the Southeast region, and the Southwest region.
The population covered by the 1999 DHS is defined as the universe of all women age 10-49 who were either permanent residents of the households in the 1999 NDHS sample or visitors present in the household on the night before the survey were eligible to be interviewed. In addition, in a subsample of one-third of all households selected for the survey, all men age 15-64 were eligible to be interviewed if they were either permanent residents or visitors present in the household on the night before the survey.
Name |
---|
National Population Commission |
Name | Role |
---|---|
ORC/Macro | Technical assistance |
Name | Role |
---|---|
United Nations Population Fund | Funding |
U.S. Agency for International Development | Funding |
The 1999 Nigeria Demographic and Health Survey (NDHS) is a nationally representative probability sample of women age 10-49 living in households. The sampling frame used for the survey was constructed from the enumeration areas (EAs) into which the country was delineated for the 1991 population census. Currently, the frame contains 212,079 EAs.
The sample was stratified into rural and urban areas and was selected in two stages. It was designed to produce reliable estimates of most of the variables for the rural and urban segments of the country as well as each of five statistical regions, namely, the Northeast region, the Northwest region, the Central region, the Southeast region, and the Southwest region. Each of these five regions was treated as a sampling domain. The distribution of the states across these regions is shown fully in Appendix A. The regions used for this survey differ from the six geopolitical zones of the country and the seven administrative zones of the National Population Commission.
The primary sampling unit was the EA. Altogether, 400 EAs were selected with equal probability. In all, 119 urban EAs and 281 rural EAs were selected. To ensure data quality, the selection of the EAs was done centrally by trained statisticians at the Liaison Office of the National Population Commission (NPC) in Lagos. The list of selected EAs was sent to the NPC offices in each state to identify the EAs, draw sketch maps, and conduct a listing of all households in each selected EA. NPC's comptrollers at the local government offices thereafter cross-checked the work of the state officers to ensure no omission of any building within the EA.
At the second sampling stage, one in every five households listed was selected for interview. The combination of equal probability selection at the first stage and a fixed sampling rate at the second stage yielded a roughly self-weighting sample design. However, while the returns from the rural stratum showed an appreciable level of self-weighting, the returns from the urban stratum showed a significant level of deviation from self-weighting. The deviation in the urban stratum was due to under listing of dwellings in some EAs because of changes in EA boundaries over time. Therefore, in processing and estimating the population parameters, the sample returns were weighted by considering the selection probabilities of the primary sampling units, the expected and eventual field returns, and the differential response rate among the domains. The weights were standardised and entered with the individual data records. Thus, all the tables presented in this report are based on weighted data.
In the selected households, all women age 10-49 were eligible for interview with the Women's Questionnaire. In every third household, men age 15-64 were eligible for interview with the Men's Questionnaire.
A total of 7,919 households were sampled, of which 7,736 were determined in the field to be valid households and 7,647 were successfully interviewed, giving a response rate of 99 percent.
Of the 8,918 eligible women age 15-49 in these households, 8,199 were interviewed for a response rate of 92 percent. Every third household was selected for coverage with the Men's Questionnaire. Thus, 2,620 households were sampled, of which 2,571 were found and 2,550 were successfully interviewed. In these households, a total of 3,082 men age 15-64 were identified and 2,680 were interviewed for a response rate of 87 percent.
Four questionnaires were used for the main fieldwork: the Service Availability Questionnaire, the Household Questionnaire, the Women's Questionnaire, and the Men's Questionnaire.
a) The Service Availability Questionnaire was implemented at an early stage of the fieldwork and was designed to assess the availability (or supply) of health and family planning services. It was administered at the community level (enumeration area) by interviewing knowledgeable informants in the selected community.
b) The household questionnaire was used to identify both men and women who were eligible for the individual questionnaire and to collect data on housing characteristics. name, sex, age, and education.
c) The Women's Questionnaire was administered to all women age 10-49 who were listed on the Household Questionnaire. The decision to interview women age 10-14 was influenced by pretest findings on teenage pregnancy, motherhood, and the age at commencement of sexual activities. Since most of the variables presented in this report are not relevant for the youngest women, the analysis has been restricted to women age 15-49. Women were asked questions on the following topics:
d) The Men's Questionnaire was used to interview men age 15-64 living in every third household. It was similar to that for women except that it omitted the sections on antenatal and delivery care, breastfeeding, vaccinations, causes of death, female genital cutting, and height and weight.
Start | End |
---|---|
1999-03 | 1999-05 |
Name |
---|
National Population Commission |
The NPC Comptroller of the local government who is a very senior staff thereafter cross checked the work of the technical staff to ensure no omission of any building within the EA or inclusion of a building outside the boundaries of the EA. After approval of the building numbering and listing, the technical staff who did not serve as interviewers in the EA identified and listed all households within the EA in the Household Listing Form -NDHS-07. The Comptroller again was supposed to spot check the listed households by re-listing all households in one of five residential buildings listed by the technical staff.
The following quality control procedure was adopted:
i) If no error was found in the re-listing (sample), then the listing was accepted for enumeration,
ii) if 2 or more percent error was found then, the entire EA was re-listed
iii) if errors were found but less than 2 percent, a second independent sample was drawn, the cumulative errors were found, if 2 or more percent error was obtained (from the two samples), the entire households in the EA will be re-listed, otherwise correction was to be made on the Household Listing Form (NDHS-07). (Note that in an EA, where there are less than 10 residential buildings the comptroller is expected to quality check one of every five households listed in the EA).
TRAINING
Two levels of training were organised. The first level was the training of trainers, which took place in Lagos between 16 and 20 November 1998. The trainees consisted of zonal and state directors of NPC and selected senior headquarters/liaison office staff who are well versed in survey methodology. Individuals who participated at some of the workshops organised at the planning stages of the survey acted as the facilitators during this level of training. The second stage of training took place for two weeks at the seven zonal headquarters of the NPC (namely, Kano, Yola, Port Harcourt, Enugu, Lagos, Ibadan, and Kaduna.) This level of training involved the training of interviewers, supervisors and field editors. Those trained at the first level of training facilitated at this level.
FIELDWORK
Immediately after the training exercise, NDHS field personnel went to the field for data collection. The field staff consisted of 34 teams, each composed of one supervisor, one field editor, four female interviewers, one male interviewer, and a driver. Fieldwork was carried out in 400 EAs nationwide between 29 March 29 and 29 May 1999.
The personnel who took part in the processing of NDHS data consisted of 20 data entry operators, two supervisors, and six coders/editors, all of whom are staff of the NPC. Before data processing began, the data entry operators were trained intensively for two weeks by staff from Macro International Inc. (USA).
Data were processed on microcomputers and printers that were provided by Macro International Inc., with funding from USAID. The computers were used to establish the nucleus of a demographic laboratory at the NPC. Data were processed using programmes written by Macro International Inc. with the Integrated System for Survey Analysis (ISSA), which was designed for processing DHS data.
Sampling errors, on the other hand, can be evaluated statistically. The sample of respondents selected in the NDHS is only one of many samples that could have been selected from the same population, using the same design and expected size. Each of these samples would yield results that differ somewhat from the results of the actual sample selected. Sampling errors are a measure of the variability between all possible samples. Although the degree of variability is not known exactly, it can be estimated from the survey results.
A sampling error is usually measured in terms of the standard error for a particular statistic (mean, percentage, etc.), which is the square root of the variance. The standard error can be used to calculate confidence intervals within which the true value for the population can reasonably be assumed to fall. For example, for any given statistic calculated from a sample survey, the value of that statistic will fall within a range of plus or minus two times the standard error of that statistic in 95 percent of all possible samples of identical size and design.
If the sample of respondents had been selected as a simple random sample, it would have been possible to use straightforward formulas for calculating sampling errors. However, the NDHS sample is the result of a two-stage stratified design, and, consequently, it was necessary to use more complex formulae. The computer software used to calculate sampling errors for the NDHS is the ISSA Sampling Error Module. This module used the Taylor linearisation method of variance estimation for survey estimates that are means or proportions. The Jackknife repeated replication method is used for variance estimation of more complex statistics such as fertility and mortality rates.
The Jackknife repeated replication method derives estimates of complex rates from each of several replications of the parent sample, and calculates standard errors for these estimates using simple formulae. Each replication considers all but one clusters in the calculation of the estimates. Pseudo-independent replications are thus created. In the NDHS, there were 399 non-empty clusters. Hence, 399 replications were created.
In addition to the standard error, ISSA computes the design effect (DEFT) for each estimate, which is defined as the ratio between the standard error using the given sample design and the standard error that would result if a simple random sample had been used. A DEbT value of 1.0 indicates that the sample design is as efficient as a simple random sample, while a value greater than 1.0 indicates the increase in the sampling error due to the use of a more complex and less statistically efficient design. ISSA also computes the relative error and confidence limits for the estimates.
Sampling errors for the NDHS are calculated for selected variables considered to be of primary interest. The results are presented in an appendix to the Final Report for the country as a whole, for urban and rural areas, and for the five regions. For each variable, the type of statistic (mean, proportion, or rate) and the base population are given in Table B.1 of the Final report. Tables B.2 to B.9 present the value of the statistic (R), its standard error (SE), the number of unweighted (N) and weighted (WN) cases, the design effect (DEFT), the relative standard error (SE/R), and the 95 percent confidence limits (R_+2SE), for each variable. The DEFT is considered undefined when the standard error considering simple random sample is zero (when the estimate is close to 0 or 1). In the case of the total fertility rate, the number of unweighted cases is not relevant since there is no known unweighted value for woman-years of exposure to childbearing.
The confidence interval (e.g., as calculated for children ever born to women aged 15-49) can be interpreted as follows: the overall average from the national sample is 2.848 and its standard error is .04. Therefore, to obtain the 95 percent confidence limits, one adds and subtracts twice the standard error to the sample estimate, i.e., 3.848-+2x.04. There is a high probability (95 percent) that the true average number of children ever bona to all women aged 15 to 49 is between 2.771 and 2.925.
Sampling errors are analysed for the national sample and for two separate groups of estimates:
(1 ) means and proportions, and (2) complex demographic rates. The relative standard errors (SE/R) for the means and proportions range between 0 percent and 50.7 percent with an average of 6.6 percent; the highest relative standard errors are for estimates of very low values (e.g., currently using implants among currently married women who were currently using a contraceptive method). If estimates of very low values (less than 10 percent) were removed, then the average drops considerably. So in general, the relative standard error for most estimates for the country as a whole is small, except for estimates of very small proportions. The relative standard error for the total fertility rate is small, 2.2 percent. However, for the mortality rates, the average relative standard errors are somewhat higher, e.g., 4.8 percent for under-five mortality.
There are differentials in the relative standard error for the estimates of sub-populations. For example, for the variable with secondary education or higher, the relative standard errors as a percent of the estimated mean for the whole country, for the rural areas, and for the Northeast region are 3. I percent, 4.4 percent, and 17.5 percent, respectively.
For the total sample, the value of the design effect (DEFT) averaged over all variables is 1.46, which means that due to multi-stage clustering of the sample variance is increased by a factor of 1.46 over that in an equivalent simple random sample.
Any assessment of the quality of survey-based data will find internal and external inconsistencies. Sampling variability can contribute to such findings especially when considering data at the regional as opposed to the national level. So, a data quality assessment often requires a judgement as to whether the degree of the inconsistency indicates acceptable departures from expected patterns or severe data problems.
There are a number of problems with the data of the 1999 NDHS. There is clear evidence of underreporting of events for the time period 1984-89. The magnitude of the mortality decline implied by the estimates from the 1990 and 1999 surveys ranks among the largest observed in high-mortality African countries. Yet, the health indicators for Nigeria indicate a deterioration of immunisation coverage for children over the last decade. The neonatal mortality rate for the Northeast Region is unrealistically low and inconsistent with the postneonatal mortality rate. Both the mortality and the fertility data for the Central Region appear particularly flawed.
The weight of evidence indicates that the mortality rates based on the data are most probably underestimates. Moreover, the nature and scope of the data defects leading to this conclusion suggest that the possibility of repairing these data so that they would form the basis for reliable mortality estimates for Nigeria is not good. This review is useful because of the implications for future surveys that attempt to estimate mortality in the Nigerian setting. It is not the purpose here to specify the design parameters that are necessary to ensure that reliable data are collected. Such design features are well known and it should be a high priority of the next survey to put them in place.
Name | Affiliation | URL | |
---|---|---|---|
MEASURE DHS | ICF International | www.measuredhs.com | archive@measuredhs.com |
Use of the dataset must be acknowledged using a citation which would include:
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.
Name | URL | |
---|---|---|
General Inquiries | info@measuredhs.com | www.measuredhs.com |
Data and Data Related Resources | archive@measuredhs.com | www.measuredhs.com |
DDI_NGA_1999_DHS_v01_M
Name | Role |
---|---|
World Bank, Development Economics Data Group | Generation of DDI documentation |
2012-05-03
This site uses cookies to optimize functionality and give you the best possible experience. If you continue to navigate this website beyond this page, cookies will be placed on your browser. To learn more about cookies, click here.