Probability sampling offers the advantages of less biased results and a higher representation of the sample in question. It also allows for accurate statistical inferences to be made.
Probability sampling is useful for studying units of both similar and different samples within a group. Random types of probability sampling allow for the elimination of any possible conscious or inherent bias in those conducting the study as the samples are selected at random. Focusing on the features and behavior of the sample in relation to the larger group they are a part of is called statistical inference, and helps generalize the overall behavior of a group. Statistical data on a sample may be collected through direct observation or surveys.
Simple random and systemic sampling are both forms of probability sampling that focus on similar traits of individuals within a unit, while stratified random sampling focuses on individuals in a group that exhibit a variety of different traits. Cluster sampling is used to study the behavior of units within a group rather than individuals, and is less accurate than individual-based types of probability sampling. Probability sampling is useful in studies where full representation of a group is desired, as opposed to less focused types of sampling, such as convenience or quota sampling.