If the population of a study is finite and just covers only a small set of data and the time and cost of the study is of no concern to the one who is conducting the study then it is better to take samples rather than conduct the study over the whole population