Different Sampling Methods
[Soben Peter 4th ed 368-69/ 5th ed 591]
Different sampling methods depending upon the type and nature of the population and the objectives of the investigation are:
1) Non probability sampling:-
Non-probability samples are not truly representative and are therefore less desirable than probability samples. This method is used in cases where a researcher may not be able to obtain a random or stratified sample, or it may be too expensive or when it may not be necessary to generalize to a larger population. The three types are:
a) Quota sampling
Here the general composition of the sample is decided in advance and includes a particular segment of population.
Eg: If a researcher is interested in attitudes of members of different states, he could set a quota of 3% quota people of each state. However, the sample may no longer be representative of actual proportions in the population.
b) Purposive sampling:
Purposive sampling is done to save a very specific need or a purpose. A researcher may have a specific group in mind, such as high Level business executives.
A subset of purposive sample is a 'Snow ball" sample/ chain referred sample" in which the researcher picks up the sample along the way, analogous to a snowball accumulating snow. A snow ball sample is achieved by asking a participant to suggest someone else who might be willing or appropriate for the study. Snow ball samples are particularly useful in hard to track populations, such as those with illegal behavior like drug users, homeless people etc.
c) Convenience sampling:
A convenience sample is a matter of taking what you can get. It is not randomly obtained volunteers would constitute a convenience sample.
In this method, the sample selection depends exclusively on the judgment of the investigator.
2. Probability sampling:
It is the recommended method of sampling in which each individual unit in the population has a known probability of being selected. It includes:
a) Simple random sampling
- Lottery method
- Table of random numbers
a) Stratified sampling
- Stratified random sampling
b) Cluster sampling
3. Other sampling methods
a) Multiphase sampling
b) Multistage sampling
Simple Random Sampling
i) It is applicable when the population is small, homogenous and readily available.
ii) Used for clinical trials like testing the efficacy of a particular drug.
Eg: Ward patients, patients coming to hospital.
i) It is applicable when the population is Large, non- homogenous and scattered.
ii) Sample is calculated by taking every kth variable, where 'k' is
Sample size desired
i) It is used when population is large and not homogenous
ii) Initially, the population is divided into homogenous subgroups known as strata and a sample is drawn randomly from each stratum.
Stratified Random Sampling:
i) In this technique the population to be sampled is subdivided into groups known as strata, such that each group is homogenous in its characteristic.
A simple random sample is then chosen from each stratum. This type of sampling is used when the population is heterogeneous with regard to the characteristic under study. This method ensures more representativeness, provides greater accuracy and car concentrate on a wider geographical area.
i) Cluster sampling involves grouping the population and then surveying the entire population
ii) It is used when units of population are natural groups or dusters such as villages, schools, wards, blocks, etc.
i) Here sampling is done in different phases. For example: in tuberculosis survey.
* First phase - Montoux test is done in all cases of sample
* Second phase - X-ray of chest taken in Montoux positive cases.
* Third phase - Sputum examination in X-ray positive patients.
ii) Advantages - Less cost, less laborious and more purposeful.
i) It is employed in large country surveys.
ii) Sample procedures carried out in several stages using random sampling techniques.
Country wide survey
District wide survey by selecting some districts randomly
Village wide survey by selecting some villages randomly
Some units are selected randomly