In random sampling, the sample is drawn according to prespecified chances from the population, and thus it is also called probability sampling. Since planned randomness is built into the sampling design according to the probabilities, one can use these probabilities to make inferences about the population. For example, if one uses the sample mean to estimate the population mean, it is important to know how the sample is being drawn since the inference procedures such as confidence intervals will depend on the sampling scheme. Similarly, hypothesis testing on the population mean also depends on the sampling scheme used.

One important purpose of random sampling is to draw inferences about the population. On the other hand, if the sampling scheme is nonrandom, that is, not all outcomes ...