The process which involves the selection of the sample but does not involve its composition is called as the sampling method. There are different methods for carrying out the sampling process and these methods can be categorized as follows –
a. Probability Sampling – Studying different methods for the sampling is very important as its study helps in the improvement of the precision, accuracy, efficiency etc. of the different sample results, as these are very much dependent on the actual method that is being used.
This method of sampling, as the name suggests, depends on the theory of the probability. The selection of the units from a population is done depending on the probabilities that are known. Probability sampling is a process of sample selection in which the elements are chosen by chance methods through tables of the random numbers.
The probability sampling method is also referred to as the random sampling method and is mainly used in the cases involving significant decisions, not giving much importance to the budget and the time.
Probability sampling methods are of the following types –
1. Simple Random Sampling –
• Is free from the personal bias.
• Sample is drawn in such a way that the every member of the population has an equal chance of being included in the sample.
• Sample obtained is referred to as the Random Sample.
• Very suitable, if the population is small or if the list of the elements in the population can be made.
• Very useful in the cases involving a homogeneous population.
• Is further of two types – simple random sampling with replacement and the other one is the simple random sampling without replacement.
• If the units of a sample are drawn one by one from the population in such a way that after every drawing the selected unit is returned to the population then this is called as the simple random sampling with replacement.
• And if the units of a sample are drawn one by one from the population in such a way that after every drawing the selected unit is not returned to the population then this is called as the simple random sampling without replacement.
Random sampling can be performed with the help of certain specific methods and some of these methods are –
A. Lottery method –
• A lottery is drawn by writing a number or the names of various units and then putting them in a container.
• They are completely mixed and then certain numbers are picked up from the container.
• Those picked are taken up for the sampling.
B. Tippet’s Numbers method –
• Was evolved by L.H.C Tippet, name of this method comes from his name.
• He made a list of 10, 4000 four digit numbers written at random on each page.
• From these numbers, samples are drawn at random.
• Is a very reliable and a dependable method.
C. Selection from Sequential List –
• Names are arranged serially in a specific pattern.
• Pattern may be alphabetical, geographical or it may be serial in nature.
• Then from the list, any number may be taken up.
• Selection can be started from anywhere.
D. Grid System –
• Involves drawing of the map of the entire area.
• Then screen with the squares placed on the map of the entire area is drawn.
• After this the screen with the squares placed upon the map and some of the squares are selected at random.
• Then the screen is placed upon the map and the area falling in the selected squares is taken as the sample.
2. Stratified random sampling – This type of probability sampling method is one of the most commonly used methods and involves the division of the whole population into a number of strata.
These strata are very much exclusive and also very exhaustive in the nature. From each of these strata, a simple random sample is drawn, by this; the number of the samples drawn from each of the sample becomes proportional to their respective strata size. When the population is heterogeneous in nature, this type of sampling plays a very critical role. The stratification in this method is performed in such an elegant way that the variance between the strata is high and data within each stratum is very small.
Types of Stratified Random Sampling –
A. Disproportionate Stratified Random Sampling – In this type of sampling, equal number of the units are drawn from each stratum, not depending on the size of the strata.
B. Proportionate Stratified Random Sampling – in this type of sampling, the number of units in each stratum is proportionate to its number in the universe.
3. Systematic random sampling – This method involves the formation of a sample in a very systematic manner, involves the arrangement of the units in the population in a serial manner. A major point to be kept in mind here is that the population should be finite in nature and also should be defined very clearly.
After this, from the first K units, one unit is selected at random, this unit and also every K th unit onwards from the serially listed populations forms a systematic sample. This method is very simple and convenient in use as it saves a lot of time.
4. Multi Stage sampling – As the name suggests, this method takes place in a number of stages population consists of first stage units and each of these first stage units consist of a number of second stage units. These second stage units further consist of the third stage units and this is followed till the ultimate unit is achieved.
One of the major positive points of such a sampling method is that the frame of the second stage units is needed only for the selected first stage units. In this sampling procedure, sample from the first stage units is chosen with the help of any method that is suitable or correct. After this the sample of the second stage is chosen by suitable method from first stage units and then this procedure is followed until the ultimate units are reached.
5. Cluster Sampling – This type of method is very useful in lowering the filed cost as this method is a very practical and easily opera table method. Here the population is divided into a number of groups and these groups are referred to as the clusters and further these clusters are referred to as the primary sampling units. This method involves first of all the identification of the cluster, according to which clusters may be units like districts, talukas, city blocks, schools etc. Clusters should be homogeneous in the internal characteristics.Then further in this method, determination of the number of the stages is to be done.
It may involve single stage or two stage or multi stage sampling.
b. Non Probability Sampling – This type of method does not provide any type of ground for estimating the probability for each item in the population process to be included in the sample. Here sampling error is not measurable and such methods are used widely to contact the respondents in this case as the primary objective involves the probing for the possible range of the answers.
Non Probability Sampling methods are mainly classified as –
1. Convenience sampling –
• Selection of the samples is done according to the convenience of the researcher.
• Representatives of the samples cannot be known.
• So as a result of this, results that are received are biased in nature.
• Not suitable for use in the descriptive and the casual studies.
• Suitable for the exploratory studies.
• Very useful in testing of the questionnaire.
2. Judgment Sampling –
• Also called as the Purposive sampling.
• Firstly, a sample is drawn from the population, which a researcher thinks to be a representative of the population.
• All the members do not get the chance to get selected in the sample.
3. Quota Sampling –
• Form of the Stratified sampling.
• Number to be selected from each stratum is referred to as the Quota.
• Quota is selected in advance.
• Used in the marketing research studies.
• Sample is selected on the basis of parameters like the age, sex, income etc.
• Field workers mainly choose the sample and are assigned quotas.