Sampling
Inferential Statistics is that branch of statistics in which samples taken from populations are use to make inferences about the population. The techniques of inferential sataistics assume that all objects or persons in a population have an equal chance of being selected into the sample. Samples meeting this dictum are called unbiased samples. The researcher has several sampling methods available to ensure an unbiased sample.
- Random samples are selected by assigning a unique number to every person in a population. Then have a computer randomly select numbers from the list until the desired sample size has been reached. Another method is to place all numbers in a hat , mix them well, and then randomly select numbers from the hat until the desired sample size is reached.
- Systematic Samples are selected by assigning a unique number to every person in a population and then selecting every third number, or every other number, etc.
- Stratified samples are selected by partitioning the population into unique groups based on some characteristic such as sex, educational level, income level etc. Then either a random or systematic sample is taken from each group.
- Cluster samples are selected by using groups that already exist rather than creating them. Suppose a study involving 3rd graders in a large city school system was planned. Suppose further that there were 75 elementary schools in the system together with 10 to 20 third grade classes per school. Rather than attempting to get a random sample of all 3rd graders in the system, a random sample of say 15 of the schools themselves could be selected for the study.