Q.Distinguish between a census and a sample survey. Give one business situation where a census would be necessary and one where a sample survey would be preferred, giving reasons.

A census and a sample survey differ mainly in coverage - the whole population versus only a part of it - and the choice between them in business is usually driven by cost, time and whether complete accuracy is legally or practically required.

Census (complete enumeration): every single unit of the population is examined; it gives, in principle, the most complete and accurate result but is costly, time-consuming, and sometimes physically impossible.

Sample survey: only a part (a sample) of the population is examined and the results are generalised to the whole population; it is quicker and cheaper, and is the only option when testing an item destroys it or the population is too large to enumerate fully within a reasonable time and budget.

Situation needing a census: A statutory financial audit of a company's books of account - the law requires every transaction to be verified, so no sampling is permissible; a sample survey would not satisfy the legal/audit requirement of complete verification.

Situation preferring a sample survey: Testing the average life (in burning hours) of a batch of 10,000 light bulbs manufactured in a day. Since testing a bulb to failure destroys it, a census (testing all 10,000) would destroy the entire batch and leave nothing to sell - a random sample of, say, 50 bulbs is tested instead, and the result is generalised to the whole batch.

Students sometimes assume a census is always 'better' because it is complete - in destructive testing or very large/dispersed populations, a census may be impossible or pointless, and a well-designed sample is the only workable choice.