Sampling refers to the investigation of a part of the whole population or universe. A sampling procedure is a technique of selecting a sample from a given population. A statistical sample, according to Calvin, is a miniature picture or cross-section of the entire group or aggregate from which the sample is taken.
The entire group from which a sample is chosen is known as the “population”, “Universe”, or “supply”. In short, sample represents the whole population and by observing the sample, certain inferences may be made about the population. For collecting representative data, samples are not selected haphazardly but a proper procedure is adopted, so that the influence of chance and probability can be estimated.
The important consideration in selecting a sample is to see that it is a close representative of the universe. The size of the sample may not be a guarantee of its being representative of the population. Sometimes a large sample poorly selected may not prove to be a true representation of the universe while a small sample properly selected may be much more reliable. The actual selection of a sample should be so done that every item in the universe under must have the same chance for inclusion in the sample.
An adequate sample is one that contains enough cases to ensure reliable results. But question arises, ‘How large a sample should be?’ Most often, we hear such answers as ‘Never less than 100′ or’500′ or’5 per cent of the population.’ However, the answer to above question can be given only when we have at our disposal (i) the designation of the parameters which one wishes to estimate, (ii) the range of unreliability permissible in estimates, and (iii) a rough estimate of the dispersion of the investigated characteristic. We can therefore, say that the rule of thumb, answers of arbitrary numbers or percentages to the above question are misleading, and one should insist on being given the information required, even if in the form of extremely rough estimates, before attempting to answer the question.