A sample is the specified subset of the population selected for observation and analysis in research. As population often times is too large to observe, sample provides a more specific and narrower subject to study. There are two available models of sampling methods that can be used in research which are probability and non-probability sampling.
1. Probability Sampling
To ensure that no systematic bias occurs in the selection of elements, probability sampling adopts a random selection procedure. It is the commonly used model in quantitative research and by its statistical inference considered to be the representative of reality. There are four common types of probability sampling used to determine the number of the research samples named simple random sampling, systematic sampling, stratified random sampling, and cluster sampling.
- Simple Random Sampling
Simple random sampling is the easiest and simplest method out of all four types. Every member of the population has an equal chance of being selected as a sample in this method. Like its name, simple random sampling randomly selects the assigned number of samples needed for the research, without prior steps to differentiate the population into groups. It usually utilizes technology, such as a random number generator, and is an ideal method to draw a sample for research. In spite of this, if a large population is involved, using simple random sampling could be quite time-consuming and challenging.
- Systematic Sampling
The variant of simple random sampling is systematic sampling. This method selected every nth member from the research population or sampling frame. To determine the sample, first researcher will have to identify the population and determine the interval for every nth dividing the population with desired sample size. Then, the researcher will have to randomly select the first member and selecting every nth member until the sample size is achieved.
- Stratified Random Sampling
Through stratified random sampling, the research population is first divided into subgroups or strata. Then, a random sample from each subgroup will be selected through either simple random sampling or systematic sampling method. The stratified random sampling has its own advantage and is suitable for research that has large populations (because the population and key subgroups are accurately predicted). Researchers are able to determine the criteria for each stratification by themselves based on their research questions or hypothesis.
- Cluster Sampling
Similar to stratified random sampling, cluster sampling selected the research samples based on subgroups of the population. However, what differentiates this from stratified random sampling is the subgroups of the population are not created by the researcher, but a naturally occurring cluster. This method is the most suitable to apply in research that has a large population scattered across a wide geographical area.
