Types of Sampling
Some Vocabulary Homogeneous groups: All members of the group have a characteristic that is the same. Heterogeneous groups: all members of the group have characteristics that differ but make up the characteristics of the entire population
Stratified sampling Split the population into homogeneous groups before selecting a sample. Then use simple random sampling within each strata to make a larger sample.
Stratified sampling Example Split the school into grades (each grade is a strata) Do a simple random sample to choose 10 people from each grade. Combine the 10 people from each grade to make a random sample of 40 people from the school.
Cluster Samples Split the population into heterogeneous groups (clusters). Randomly select a cluster (or a few clusters) to make up your sample of the population. Take a Census of that cluster.
Cluster Sampling Example When studying ages of doctors in Delaware… Each hospital is a cluster. Select a hospital as your sample and survey ALL of the doctors in that hospital.
Stratified vs. Cluster Think of a Boston cream pie which consists of a layer of yellow cake, a layer of crème, and a layer of another cake, and then chocolate frosting…
Layered Cake Cluster: Taking a vertical slice of the cake. Learn about the whole pie and all of its layers combined by taking that vertical slice Strata: take a random piece of yellow cake, random piece of chocolate, random piece of crème. Get an idea of the entire cake from parts of each horizontal layer
Systematic Sampling Put all people in the population in a random order and then select every n-th member. Careful: Order of the list must not be grouped in any way related to the study.
Systematic example Put the school in alphabetical order and select every 20 th name.