The four types of probability samplings
Simple random sampling: a technique where samples are randomly selected from the population.
Systematic sampling: a means of selecting a sample where the researcher selects the nth subject from a complete list of potential participants.
Stratified random sampling: a methodology where the population is divided into mutually exclusive categories and the subjects of study selected using simple random.
Cluster random sampling: it entails dividing the population into heterogeneous clusters and using a simple random strategy to identify subjects.
Probability sampling is preferred because it minimizes the presence of systematic error and bias. These techniques produce highly representative samples of the population. As a result, they increase the reliability of the research findings because the subjects represent the characteristics of the population. Besides, probability sampling is desired because it enhances the accuracy of error estimation and provides a convenient means of making inferences about the population.
Over-sampling refer to a technique designed to increase the frequency of subjects with low prevalence in a sample. This technique may be applicable in circumstances where random sampling results into a few members of a certain community, which may not provide representative traits about the population. It is used to solve the problem of imbalanced classification and increasing the prevalence of minority class.
The margin of error refers to the radius of error expected from survey results. It is a measure of assurance and expresses the confidence interval in a poll. The margin of error is expressed in percentage form, and a larger error denotes less confidence in survey polls.
The four types of non-probability samplings methods are;
Judgment
Snowball
Quota