The law of large number (as well as the theory of large sampling) implies that as the number of observations increases, the value of the estimator of a parameter based on the observations becomes closer to the true value of the parameter. In view of this theory, complete census should be better than the sample survey. But still sample survey is considered to be better than complete census. Is it not a contradiction ?

More Dhritikesh Chakrabarty's questions See All
Similar questions and discussions