Chat with our AI personalities
stratified sampling, in which the population is divided into classes, and random samples are taken from each class;cluster sampling, in which a unit of the sample is a group such as a household; andsystematic sampling, which refers to samples chosen by any system other than random selection.
In some situations stratified random sampling may be more appropriate. You may have a population which can be divided up into a number of subsets (strata) such that the difference between units in different strata is much greater than the difference between units within each stratum. A probability sample may not have enough units from some of the smaller strata. A stratified random sample will ensure that each stratum is represented proportionally. In other situations, cluster sampling may be more appropriate. Suppose you wish to visit a sample 1% of all schools in the country. If you were to choose the schools by probability sampling they would be all over the country and you would require a huge amount of time and money to visit them all. What you could do, instead, is to divide up the country into 1000 regions. Select 10 of these regions (1%) and then visit every school in the selected regions. Far less running around!
A sampling distribution refers to the distribution from which data relating to a population follows. Information about the sampling distribution plus other information about the population can be inferred by appropriate analysis of samples taken from a distribution.
Convenience sampling is also know as grab sampling. There is no procedure for the sampling itself because the emphasis at this stage is usually on improving other aspects of the research such as exposing flaws in a survey form or training personnel. In grab sampling you simply take any sample element that you can find although you might favour those that would exercise parts of your system that might seem weak. For instance, if your survey instrument asks for ages and some people were reluctant to provide them, then how would this be resolved once the grab sampling phase had been completed and actual sampling had started?
Random sampling is a method of selecting a sample where each member of the population has the same probability of being included in the sample. An equivalent statement is that each subset of the population, of the given size, has the same probability of being selected as any other subset of that size.