M. Com

What is Sampling? Discuss various types of Sampling.

What is Sampling? Discuss various types of Sampling.

What is Sampling? Discuss various types of Sampling.

What is Sampling? Discuss various types of Sampling.

Or

Bring out clearly the important features of (a) Simple Random Sampling (b) Stratified Sampling (c) Systematic Sampling.

Ans.

Sampling

The process of selecting a sample from a population is called sampling. In sampling, a representative sample or portion of elements of a population or process is selected and then analysed. Based on sample results, called sample statistics, statistical inferences are made about the population characteristic. A doctor examines a few drops of blood to draw conclusions about the nature of disease or blood constitution of the whole body.

Methods of Sampling

The sampling methods may be broadly classified as follows:

(A) Random Sampling

Random sampling is also called as ‘Chance Sampling’ or ‘Probability Sampling’. In this method of sampling all units of the sample are selected by chance because each unit of the universe has an equal chance of being included in the sample. The main types of random sampling are as follows:

(1) Simple or Unrestricted Random Sampling

Meaning and Definition : In practice simple random sampling is called random sampling only. It refers to that sampling technique in which each and every unit of the universe has an equal chance of being included in the sample.

Merits of Random Sampling :

1. There is no possibility of personal prejudice or bias affecting the results because the selection of items in the sample depends on chance.

2. It is more representative than judgement sampling.

3. Accuracy of the results can be evaluated. Sampling error can be determined as it follows the principle of chance.

4. The theory of Random sampling is subjected to further use in other surveys.

Disadvantages of Random Sampling :

1. It may be very costly particularly where populations are geographically dispersed and individuals are difficult to trace.

2. Size of sample is larger in Random sampling as compared to stratified sampling.

3. It is sometimes difficult for the investigator to have up to date lists of all the items of the population to be sampled.

(II) Restricted Random Sampling

If random sampling is used with certain conditions or restrictions, it is called restricted random sampling. Such sampling may assume following forms:

(1) Stratified Random Sampling: If there is lack of homogeneity in units of a universe then total units are divided into a number of groups, parts or strata and from each strata certain items are selected on the basis of simple random sampling. Suppose, there are 1,000 families in a city and we have to study family budget on the basis of a sample of 20 families. In this case we can divide 1,000 familites in certain groups on the basis of income of families. Suppose, four groups are constituted-income upto Rs. 1,000 per month, Rs. 1,001 to Rs. 3,000, Rs. 3,001 to Rs. 5,000 and above Rs. 5,000 per month. After this grouping, 20 familites will be selected in all on the basis of random sampling from these four groups.

It is worth-mentioning that stratified sampling may also be called as ‘Mixed Sampling’, because there is mixture of deliberate sampling and simple random sampling in this method.

Merits of Stratified Random Sampling:

(1) More Representative: If the system of stratified sampling is properly designed the sample drawn on this basis proves to be more representative.

(2) Administrative Convenience: There is an advantage of administrative convenience if stratified sampling is adopted in field surveys because the units from the different strata may be selected in such a way that all of them are localised in one or more geographical areas.

(3) Suitable in Heterogenous Units: If there is heterogeneity in nature and characteristics of different units of the population, a properly stratified sample may give more reliable results than a simple random sample of the same size.

(4) Difference in Sampling Problems: It is possible that the sampling problems may differ significantly in different strata of the population. In such a situation, stratified sampling can tackle the problem effectively.

Limitations of Stratified Random Sampling

(1) Problem of strata formation: If the population is not properly stratified, the results will be biased. Sometimes the problem of overlapping may also arise in strata formation.

(2) Problem in disproportionate sampling: Disproportional stratified sampling requires the assignments of weighs to different strata. If the weights assigned are faulty the sample will give biased results.

(3) Systematic Random Sampling: In this form of sampling, first of all the units of universe are arranged in some systematic frame such as alphabetical, numerical, geographical or chronological order, etc., and then units of sample are selected as per requirement. Here it is important that the first unit to be included in the sample is selected at random from the universe and subsequent units are selected in a definite sequence at equal spacing from one to another. Suppose, there are fifty students in a class and a sample of five students is to be selected. First of all, students will be arranged on the basis of roll number from 1 to 50. After that space interval will be decided on the basis of total units of universe divided by the sample size i.e., 50/5 = 10. The first unit of sample will be selected at random from the first 10 roll nos. Suppose, this number is 4 then the roll number to be included in the sample will be 4, (4+ 10) 14, (4+20) 24, (4 + 30) 34 and (4 +40) 44. The first unit selected at random is called as ‘Sample start’, which is denoted by i. The space interval is represented by ‘k’ and the sample size by n. The sequence of units to be included in the sample may be placed as follows:

i, i+k, i + 2k,………………… + (n-1) k

Systematic random sampling appears like a stratified random sampling because one unit is selected from each definite strata (space interval). It is important that systematic random sampling is very simple and economical. However, the efficiency of this method largely depends on two factors:-(a) the structure of universe should be complete and upto-date and (b) the units should be systematised in a random order i. e., list of names in telephone directory.

(4) Multi-Stage Random Sampling: In this method samples are selected at different stages and at each stage random sampling is used. Hence, it is called multi-stage random sampling. Suppose, we want to take a sample of 2,000 students from Degree Colleges of U.P. At the first stage, some universities will be selected at random. At second stage, some degree colleges will be selected at random from the universities selected at first stage. At the third stage, students will be selected at random from these colleges. It is important that at different stages different methods of random sampling may be followed. For instance, in the above three stages, the universities may be selected by stratified random sampling, colleges may be selected by simple random sampling and students by systematic random sampling.

Multi-stage random sampling provides flexibility and it can be completed with economy and administrative convenience even when the area of investigation is very large. However, there are chances of less accuracy in its results as compared to that of stratified random sampling.

(5) Cluster Sampling: In this method the total population is divided in some recognizable sub-divisions which are termed as clusters and a simple random sample is drawn from each cluster. This method is very useful in industrial production. For instance, if 1,000 locks are manufactured per day in a factory and we wish to make inspection of 10 locks intensively. The manufactured locks be divided into 10 lots of 100 locks each and the sample will be formed by selecting one lock at random from each lot.

(B) Non-Random Sampling

Non-random sampling methods are those which do not provide every item in the universe with a known chance of being included in the sample. In other words, the sampling process is, at least, partially subjective. Some of the important methods of non-random sampling are as follows:

(1) Deliberate or Purposive Sampling

In this method the investigator selects the units to be included in sample according to his own choice and requirements. Though this method is quite simple and convenient but possesses the following defects:

(1) Subjective Nature: The serious drawback of purposive sampling is that it is highly subjective because the selection of the sample items depends entirely on the personal convenience, beliefs, biases and prejudices of the investigator.

(2) Not suitable in large population: If the size of universe is large, it becomes very difficult to select representative units on the basis of purposive sampling.

(3) Difficulty in calculation of sampling error: Since the selection of units does not depend upon the principle of probability or chance, sampling error cannot be measured.

(II) Quota Sampling

Quota sampling is also a type of stratified sampling. In this method separate quota is fixed for each enumerator i.e., each enumerator is told in advance the number of the sample units he has to select from the stratum assigned to him. The quota may be fixed on the basis of some specified characteristics such as income group, sex, occupation, religion or political affiliation, etc. After the quota is fixed it is left entirely at the discretion and desire of the investigator to select units within that quota and generally random sampling is not used in this selection. For instance, if 10 investigators are appointed for the 10 areas of a city to study the views of voters before polling and each investigator is asked to interview 20 persons including 10 males and 10 females then in this case each investigator will collect information after interviewing 20 person selected by him according to his own. judgement and discretion.

Merits of Quota Sampling

(1) Mixture of stratified and purposive sampling: Quota sampling is also called as stratified purposive sampling because it enjoys the advantages of both. This method aims at making the best use of stratification without incurring high cost of probability sampling.

(2) Flexibility in selection of units: There is advantages of flexibility in this sampling and the investigator can make necessary changes in the units of sample. Suppose, the investigator has to take interview of 20 persons and 4 of the 20 persons selected refuse to respond, the investigator can select four other persons.

(3) Quite reliable results: If this method is executed by those skilled and experienced investigators who are well aware of the limitations of purposive sampling and if proper controls are managed on the investigators, quota sampling can give quite reliable results.

Demerits of Quota Sampling

(1) Use for pre-determined conclusions: The most important drawback of this sampling is that it can be used to verify pre-determined conclusions which may be against the reality.

(2) Biased: It may be biased on account of personal beliefs and prejudices of the investigators in the selection of units of sample and information collected from these units.

(3) Problem of sampling error: Since quota sampling is not based on random sampling the sampling error cannot be calculated.

Inspite of all these limitations, this technique of sampling is widely used in market survey, political surveys and survey of opinion poll.

(III) Convenience Sampling

In this method the investigator selects those units of universe which may be conveniently located and contacted e.g., selection of the sample of teachers from the list of teachers of the university or the selection of names from telephone directory, etc. This is very easy method of selecting sample, but at the same time it is unscientific, unreliable, unsystematic and opportunistic.

(IV) Extensive Sampling

This method is almost like a census and a large no. of items from universe are included in the sample. Only those units are left about which it seems difficult or impossible to collect information.

About the author

Salman Ahmad

Leave a Comment