Sampling is the selection of a specific set of information or data from a population. To conduct particular research, a researcher has to collect the dataset from various resources. The complete set of data or the area of research is known as population, but it also increases the complexity of research. Thus, for effective research with higher reliability and fewer complexities researcher take the sample that can encompass the whole population.
The sample is a part of the population, a subset which represents the whole population and the process of deriving the sample that can represent population is known as sampling. There are various techniques to perform the sampling process and derive the most suitable according to the type of research and its purpose are as follows:
The basic categorization of sampling can be done in two types:
Probability sampling is the technique in which whole the population gets an equal chance to be chosen in the sample. The equal chance of selection is provided to each set of the population; the significance of the selected sample is based upon the combined accuracy of unbiased estimations determined out of total population. The types of probability sampling are as follows:
• Simple random sampling: The simplest technique of sampling is simple random sampling when a researcher does not have any knowledge about the population and select the sample without any predomination, thus every part of the population will get an equal chance of selection.
• Stratified sampling: In stratifies sampling technique, a researcher categorizes the population in homogeneous units which is known as strata. Every stratum selected have a specific feature, which is different from others and the sample gets selected from each stratum. To categorize the whole population in different strata on the basis of their homogeneity and heterogeneity a researcher must have the knowledge of the population.
• Cluster sampling: In cluster sampling, the whole population is divided into different clusters and the whole cluster get selected as the sample.
• Multi-stage sampling: This technique of sampling is the combination of cluster and stratified sampling. In which the whole population is categorized into multiple clusters and out of various clusters, strata get selected. While selecting the stratum, the homogeneity or heterogeneity of data is determined according to the purpose of the research.
Non- probability sampling
In non- probability sampling technique, data get selected based upon the experience or ability of the researcher to select the sample. Thus, there are more chances of biases or lack of reliability of the sample. Generally, the sample does not represent the whole population. This type of technique is used when research will not use quantitative measures for the analysis. The types of non-probability sampling are as follows:
• Convenience sampling: In this technique, the sample will be selected from the available sources. This method of sampling is generally used when the sample is rarely available or is costly to collect.
• Purposive sampling: In purposive sampling technique, the sample will be selected according to the purpose of research.
• Quota sampling: The Quota sampling technique is used to select the most representable sample out of the population. The proportion and characteristics of the sample must be the same as the population, thus different subsets are selected from different sources and categories.
• Referral sampling: The referral sampling technique is used when a researcher does not have any idea or information about the population of research, thus the reference will be taken from the first population or other sources to get the data collected.