Probability Sampling Techniques

Before understanding probability sampling techniques, one must know the difference between sample and sampling. Sample is the subset of population which represents the population whereas sampling is the process of selecting subset from population. Now, there are two types of techniques for selecting the subset or sample from population which are: Probability Sampling techniques and Non-Probability Sampling techniques. Here we will only discuss probability sampling techniques. In probability sampling technique every object of the population has equal chance or probability of being selected as a part or respondent of the sample. This sampling technique is free from any kind of discrimination because probability or chance of any respondent is equal and varies from method to method of probability sampling. This method is divided into five broad methods are described below:

Simple Random Sampling Method

In Simple random sampling every respondent has equal probability of being selected. In this method the subset of population is achieved through chance but without any logic. For example if a researcher wants to select 10 students from a class of 70 he/she must select with out any logic. The researcher can use any logic such as: students having age above 20, students from first row, students scoring higher marks etc. Simple random sampling technique can be used in two ways, with replacement and without replacement of units.

Systematic Sampling Technique

Systematic Sampling Technique is a special type of random sampling technique in which first unit of sample is taken randomly from the population. The other units of sample are taken at a fixed interval. This sampling technique has few advantages such as: if first unit of sample is identified randomly than remaining respondents can be easily taken. Except first respondent all other units lacks complete randomization. The disadvantage of systematic sampling technique is that if the first respondent is identified from the end-points then sampling results will be misleading but it happens in rare cases. For-example: If production department wants to check the quality of goods than if first unit of sample is good, then the remaining goods will be considered as good.

Stratified Sampling

Stratified Sampling is statistically more efficient because it is an improvised technique above systematic and simple random sampling techniques. In this method, population is divided into a specified set known as strata. A stratum has respondents and units with homogenous attributes but the respondents of the one stratum do not have same characteristics as compared to the attributes of other stratum.  
For example: In order to measure the standard of living in any city, people are first divided into different classes such as upper class, medium class and lower class etc. People are divided into these classes on the basis of various attributes such as: income, education, job nature etc. If the same proportion of respondents is used in sampling, it is called as proportional stratified sampling otherwise it is known as disproportional stratified sampling. If the variance is zero or near to zero between two strata, then proportional strata sampling method would be used, otherwise disproportional strata would be used.

Cluster Sampling

It is sampling technique method in which population is classified into clusters. Different clusters are similar to each other but members of these clusters do not have same attributes. In this sampling technique each cluster is treated as a small population itself with entire attributes. Randomly any one cluster can be studied and all the respondents of that cluster would be taken in the study to achieve the results of the population.

Multi Stage Sampling

This is one of the most important sampling techniques which used for large population size such as country. It also leads to higher cost and takes more time to the research. The basic purpose of multiphase sampling is to reduce the sample size. This sampling method facilitates to divide population into small samples which must be practicable in both terms such as time and cost. Depending upon the reality, this method utilizes more than one stage to sample the population and different types of sampling methods are used in the particular stages.