A survey, which includes every element of the population, is known as census or the method
of complete enumeration. On the other hand, when a part of the population is studied and predictions are made about the population based on this part, it is called sampling.
In terms of accuracy of results, census is better as it studies all the units of population but this method is very time consuming, expensive and sometimes not feasible to use.
Hence, sampling is better due to following reasons
(i) Economical Sampling involves study of a fraction of population and hence, the cost involved in sampling is relatively low. Census costs are high especially in case of large population with wide coverage in terms of area.
(ii) Time Saving Huge amount of time is required to conduct a census survey, if the population size is large or spread over a wide area while sample studies do not take that much time to be
(iii) Lesser Effort As only a part of the population is studied, it entails lesser effort on the part of the investigator than that required in census.
(iv) Inappropriateness of Census In certain case, when the population is infinite or exhaustible, census cannot be done and hence, sampling is the only choice.
For example, one cannot burn all the units of coal available to know their calorific value sample is the only means of testing it.
(v) Considerable Accuracy Results from sampling may not be as accurate as in case of sampling but the level of accuracy of these results can be established through statistical tests of significance and hence, can be applied in general to the whole population, if found significant.