Sampling is an important aspect of social survey. It is the selection of relevant units of inquiry for the collection of data and should be done in a scientific manner. Therefore it is important to know how closely the measures based on sample represent the parameters and how much variation one may expect if other samples are analysed.The measures of reliability are concerned only with fluctuations due to random sampling and they have nothing to do with observational and computational errors. Whenever a measure of reliability is computed it is understood that the sample is adequate and has been selected according to a rigorously scientific procedure.
According to large sample theory the reliability of a measure such as the arithmetic mean depends upon the number of cases in the sample and the variability of the values in the sample. The reliability of a measure is related to the size of the sample. The degree of variability of cases in a sample also has an important influence on the reliability of the measures computed from the sample. If the cases in the sample show a pronounced scatter a greater chance of fluctuation in the measures would naturally be expected.