Survey ID Number
Demographic and Health Survey 1988-1989, Zimbabwe
Estimates of Sampling Error
Sampling error is a measure of the variability between all possible samples that could have been selected from the same population using the same design and size. For the entire population and for large subgroups, the ZDHS sample is sufficiently large so that the sampling error for most estimates is small. However, for small subgroups, sampling errors are larger and, thus, affect the reliability of the data.
Sampling error is usually measured in terms of the standard error for a particular statistic (mean, percentage, ratio, etc.), i.e., the square root of the variance. The standard error can be used also 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 as measured in 95 percent of all possible samples with the same design will fall within a range of plus or minus two times the standard error for that statistic.
The computations required to provide sampling errors for survey estimates which are based on complex sample designs like those used for the ZDHS survey are more complicated than those based on simple random samples. The software package CLUSTERS was used to assist in computing the sampling errors with the proper statistical methodology. The CLUSTERS program treats any percentage or average as a ratio estimate, r=y/x, where y represents the total sample value for variable y and x represents the total number of cases in the group or subgroup under consideration.
In addition to the standard errors, CLUSTERS 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 DEFT 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. CLUSTERS also computes the relative error and confidence limits for estimates.
Sampling errors are presented below for selected variables considered to be of major interest. Results are presented in the Final Report for the whole country, urban and rural areas, three broad age groups and three educationaI levels. For each variable, the type of statistic (mean, proportion) and the base population are given in B.1 of the Final Report. For each variable, Tables B.2-B.5 present the value of the statistic, its standard error, the number of unweighted and weighted cases, the design effect, the relative standard errors, and the 95 percent confidence limits.
The relative standard error for most estimates for the country as a whole is small, which means that the ZDHS results are reliable. There are some differentials in the relative standard error for the estimates by region and age groups. For example, for the variable, the proportion ever using a contraceptive method, the relative standard error as a percent of the estimated proportion for the whole country, for urban areas and for rural areas is 1,2 percent, 1,8 percent and 1,5 percent, respectively.
The confidence interval has the following interpretation. The mean number of children ever born among all women is 2,953 and its standard error is 0,045. Therefore, to obtain the upper bound of the 95 percent confidence limit, twice the standard error, i.e., 0,09, is added to the sample mean. To obtain the lower bound, the same amount is subtracted from the mean. There is a high probability (95 percent) that the true mean ideal number of children falls within the interval of 2,862 and 3,044.