Cluster Sampling (CS) is a sampling methodology that allows for simplified and more economic investigative sampling. Among the various forms of Cluster Sampling there is one that is known as Multistage Cluster Sampling (MCS), which is basically a repeated form of the traditional single-stage Cluster Sampling method (e.g., it is a cluster sampling with more than one stage). Supposing that a sample of students taking first-year English in U.S. colleges and universities were to be required, MCS would be a highly effective methodology given its simplicity and innate efficiency.
Before going into MCS and how it would be used in sampling students taking first-year English in U. S. colleges and universities, it is important to define cluster sampling. Simply stated, in cluster sampling, “Cluster, i.e., a group of population elements, constitutes the sampling unit, instead of a single element of the population” (Ahmed, 2009). Cluster Sampling is a method that is highly attractive, especially when dealing with overly large sample populations, because it allows for an economical form of sampling (by grouping sampling units into clusters). Of course, this clustering often results in less accurate results; other disadvantages include overlooked sample diversity and high standard errors.
CS consists on listing clusters in the sample population; clusters are selected and all of the sampling units are automatically selected for conducting the survey (or any other investigative method decided upon). MCS follows the same methodology; the sample population is also divided into clusters and a group of clusters is selected. However, an additional sampling (following the simple random sampling method) is conducted on the previously selected clusters. In other words, not all of the sampling units comprising the selected clusters are automatically chosen. Sampling is conducted on the clusters themselves and only a random sample from each cluster is surveyed (or investigated through any other investigative method). In a word, MCS consists on listing (cluster grouping) and sampling (Babbie, 2011).
Having defined Cluster Sampling, and having explained what Multistage Cluster Sampling (MCS) is, it is pertinent to go into the explanation of the steps that would go into selecting a multistage cluster sample of students taking first-year English in U. S. colleges and universities. First, it would be necessary to quantify the total number of colleges and universities operating in the U.S. After having totally quantified the number of colleges and universities, these should be grouped into clusters (based on their locations); clusters comprised only of colleges should be created (as well as clusters comprised only of universities). After having these clusters, it would be necessary to quantify how many relevant sampling units are within each cluster (e.g., how many students taking first-year English are enrolled in each college/university). Given that the methodology would be MCS, not all of the students would be automatically selected for the research. Instead, the numbers would be sampled using simple random sampling and this would produce the final number of sampling units for the study to be conducted.
There are different ways to indulge in population sampling. One way that allows for sampling large populations in an efficient (yet no so accurate) manner is Multistage Cluster Sampling (MCS). MCS is a method that consists on repeated Cluster Sampling (CS); there are many ways to perform the aforementioned sampling, including the one that was recommended here: simple random sampling (SRS). Even though MCS lacks in terms of accuracy, it is still a highly attractive option for performing sampling on large populations, including the one that was developed in this brief paper (the number of first-year English in U.S. colleges and universities).