One of the primary strengths of sampling is that accurate estimates of a population's characteristics can be obtained by surveying a small proportion of the population. Four sampling techniques are described here:
Simple Random Sampling
Simple random sampling is the most basic form of sampling. Every member of the population has an equal chance of being selected. This sampling process is similar to a lottery: the entire population of interest could be selected for the survey, but only a few are chosen at random. Researchers often use random-digit dialing to perform simple random sampling. In this procedure, telephone numbers are generated by a computer at random and called to identify individuals to participate in the survey.
Cluster sampling is generally used when it is geographically impossible to undertake a simple random sample. Cluster sampling requires that adjustments be made in statistical analyses. For example, in a face-to-face interview, it is difficult and expensive to survey households across the nation. Instead, researchers will randomly select geographic areas (for example, counties), then randomly select households within these areas. This creates a cluster sample, in which respondents are clustered together geographically.
Stratified samples are used when a researcher wants to ensure that there are enough respondents with certain characteristics in the sample. The researcher first identifies the people in the population who have the desired characteristics, then randomly selects a sample of them. Stratified sampling requires that adjustments be made in statistical analyses.
For example, a researcher may want to compare survey responses of African-Americans and Caucasians. To ensure that there are enough African-Americans in the survey, the researcher will first identify the African-Americans in the population and then randomly select a sample of African-Americans.
Common nonrandom sampling techniques include convenience sampling and snowball sampling. Nonrandom samples cannot be generalized to the population of interest. Consequently, it is problematic to make inferences about the population. In survey research, random, cluster, or stratified samples are preferable.