Sampling study is resorted to reduce cost in terms of money, man power, and time. Also when population study is not practicable (E.g. to study the mosquito population in an area) or not feasible (to estimate the Hb count of a patient) or when the study is destructive (to find the effect of a bomb), sample study is undertaken. There are two major categories of sampling,
- Random sampling or probability sampling
- Non –random sampling or non-probability sampling
Non-random sampling methods
These do not use the probability theory. It is not possible to estimate the standard error (SE) of the statistics. Many a time the samples become non-representative of the population. Common non-random sampling methods are three,
- Convenience sampling
Administrative convenience and easy accessibility are the criteria used to select the sample
- Judgment sampling
Sample is selected by personal judgment of the researcher
- Quota sampling
Quotas are fixed using convenience and judgment sampling principles and then, a sample is selected
Random sampling methods
Under these methods each member of the population has a known (equal or unequal) probability selection as a member of the sample.
Many a time the sample selected will be representative of the population. It is possible to estimate the standard error. Commonly used random sampling methods are
- Simple random sampling
- Stratified random sampling
- Cluster sampling
- Systematic sampling
- Multi-stage sampling
Simple random sampling
Simple random sampling can be done in two ways.
- Lottery method
- Random number table method
The sample selected may not have adequate representation of some segment of the population in case of much diverse population and can cause a problem of logistics. To overcome this, we resort to stratified random sampling.
Stratified random sampling
In this method, the population is divided into a number of sub-groups called strata (on the basis of factors such as age, gender, size and other parameters applicable). The sampling units within the strata are homogenous.
The advantages are less variance of the estimates and less cost. A problem encountered in samples selected is the non-co-operation of the selected units, especially in community studies. To mitigate this, cluster sampling method is used.
The study population consists of a number of natural groups called clusters, consisting a collection of study units. Required number of clusters are selected through stratified random sampling and all the study units contained in the clusters are studied. It is easy and much economic and recommended by World Health Organization.
If the population size is known, but the frame is unavailable and the preparation of frame is time consuming and costly, we can use the systematic sampling method
As the name indicates, the sample is selected in several stages. It is useful in large scale sample surveys and the advantages are frame is needed for only a small fraction and logistically convenient