Sampling talks about the statistical process of the selection and study of the characteristics of a small number of items from a large population of such items. The purpose is to draw inferences that are statistically applicable and valid for the entire population.
Researchers often use, two broad methods of sampling. They are:
- Non Random or Judgement Sampling
- Random or Probability Sampling
In the case of judgement sampling, the researcher selects the items that have to be drawn from the population on the basis of his or her own judgement about the general representation of these samples for the whole population. The possibility of any item getting chosen into the sample is dependent upon the judgement of the expert who is selecting the item. It is a simple technique and does not involve too much of cost as well.
In the case of Random Sampling, there is no role of judgement sampling in the sample selection. Each item has an equally important chance of getting included in the sample. When it is the case of random sampling, the researcher has to ensure to use a specific statistical process in order to make sure that there is equal probability for each item in the population. It helps to enable more reliable results with measurable margins of errors and degree of confidence.
In order to improve the cost effectiveness of the data collection, there are several variations of random sampling that are used by the researcher. There are some very common types of radom sampling methods. They are:
- Simple random Sampling
- Systematic Sampling
- Stratified Sampling
- Cluster Sampling
- Simple Random Sampling: It ensures that all the items in the population have equal probability of getting selected and equal chance in being a part of the population
- Systematic Sampling: The items are selected from the population at uniform interval and defined in terms of time, order or space.
- Stratified Sampling: The complete population is divided into homogeneous groups. After having defined the groups, from within the groups random samples are drawn independently. This kind of sampling works when within the population there is scope to have identifiable sub groups and they have distinct characteristics and differ significantly from each other.
- Cluster Sampling: In this kind of sampling, the population is divided into groups or clusters and the chosen clusters are drawn as samples. A well designed cluster sampling offers better results than the basis simple random sampling and it does not include or add on to the cost in any which way.