Describe simple random sampling, stratified random sampling, and multistage random sampling.
Simple random sampling means that a researcher chooses a sample size from a population so that each member of the population is equally likely to be chosen. That means that there is no system dictating how members are chosen. For example, if a researcher, such as a psychologist or sociologist, wants to understand what students at a certain college feel about required courses, he or she would put all the names of the students into the pool and choose names at random (perhaps using software to do so). Using the phone book is not a good way to get a random sample in today's world, as many people have unlisted numbers, only have cell phones, or do not have a stable home. In order to get a truly random sample, the researcher has to consider variables that limit or bias his or her sample.
In stratified random sampling, the population is divided into strata, or groups, before members are chosen randomly from each strata. For example, if a researcher wants to understand what first-, second-, third-, and fourth-year students think about a certain course in college and get a sample of each class, the researcher could divide the sample into groups first before selecting members at random. Researchers are often interested in getting samples of different ethnic, racial, socio-economic, and gender groups, so they can divide the population into these groups before getting a random sample of each group.
Multistage random sampling involves dividing populations into groups twice (or more) before collecting a sample. For example, if a researcher wants to understand the effect of a reading program in urban schools, he or she can first divide the population into groups of different schools and then divide each school into groups of teachers before sampling from each group. This method is often used to reduce the cost of conducting the research.