Sampling, Multistage

Steps and Uses of Multistage Sampling

MSS may begin with a selection of a large geographic region. A region might be as large as a country or a state within it. As an example, consider how MSS might be used to assess educational endeavors. Specifically, a researcher’s focus might be on the evaluation of high school students’ performance on either the ACT or SAT exams. Clusters within this high school population might be segmented by socioeconomic status, neighborhoods, or school districts. At the national level, the researcher may choose to utilize the United States. The next selection the researcher would have to make is determining which state to use for the analysis (e.g., Wisconsin). At this point, the selection of a country and a state were specifically chosen as a part of MSS. The next step would be to choose at random another smaller jurisdiction such as a county, city, and school district. The nearly 100 counties within Wisconsin should each have an equal opportunity to be chosen. Their names might be put in a hat and five of them would be chosen. For this example, assume the researcher selected Jefferson, Eau Claire, Dodge, Door, and Colombia counties. Within these counties are a number of cities and school districts; again, each should have an equal opportunity for being chosen. The opportunity for populations to be randomly chosen increases as the number of jurisdictions increase. Each school within this geographic sample has an equal chance of being selected for this project. Twenty high schools with an assumed sample size of 2,000 total students will be asked to participate in the investigation. In this case, the steps of contacting school boards and administrators to request permission to conduct the study would also need to be included. Ultimately, the data collection might consist of reviewing students’ files and specifically looking at their scores concerning standardized tests.

The scenario just described is one example of an MSS technique. One of the general goals is to obtain a large sample size with which to conduct an investigation. This process began with the selection of a country and a state within it. The next step included the random selection of counties and cities within the geographic region. Hypothetically, the researcher ended up with a sample size of 2,000 high school students. As the MSS technique is used, populations within each cluster could be analyzed. Scholars may also compare the population of one cluster with another. Using this technique might result in several implications. Scholars might predict post-secondary matriculation levels or how pedagogical strategies enable students to earn high scores.

...