I want to know that usually the census conducted at national basis are also sample based why they are call census, is their any reference based explanation
The difference is something of an historical by-product of the early modern period (beginning in the 17th century). Censuses, ostensibly, differ from survey samples in that they purport to "sample" the entire target population/population of interest. In practice, this never happens, and many organizations have increasingly changed both their methods and their terminology usage (e.g., Eurostat, WHO, US Census Bureau, the UN, etc.). However, the term is still used in the traditional sense, resulting in claims that are obviously false. For example, from Canada's 2011 Census: "The Census of Population collects demographic and linguistic information on every man, woman and child living in Canada"
Yet we do not even need to leave Canada's 2011 Census page to find:
"The objective of a census is to provide detailed information on the population at a single point in time. In this respect, one of its goals is to enumerate the entire population. Inevitably, however, some people are not counted, either because their household did not receive a census questionnaire...or because they were not included in the questionnaire completed for the household...Some people may also be missed because they have no usual residence and did not spend census night in any dwelling. In contrast, a small number of people may also be counted more than once"
Some references:
Encyclopedia of Survey Research Methods (Vols. I & II)
Nasrullah, my understanding of census is that is is a population survey and a sample survey is just that a sample survey. When one surveys the whole population, it is called census or population survey, and when one surveys a sample, it is called sample survey.
My point of view is: when we need to know about the structure of population, we do census but when we need to know about any item in detail, we need survey. There are so many censuses in the world which are sample based particularly in agriculture. Even in population censuses, many detailed questionnaires are filled on sample basis but that is called census. The major differences in my point of view are:
1. Census and survey both could be sample based but survey is always sample based.
2. Census covers more data items with less detail of every data item But survey covers less data items with more detail of every data item.
3. Sample size of a census is large enough (may be in millions at country level) while sample size of a survey is small (may be in thousands at country level).
4. Tabulations of census can be produced at the lowest administrative level (at which level sample was drawn, may be at district or lower level) while tabulations of a survey cannot be produced at the lowest level (to take a small sample, sampling cannot be done at the lowest level).