makes an important basic point. The other basic point is that the population value is not generally known precisely. So you will have to factor in the standard error of the population estimate.
Oh – and there is one situation that is dangerous : making an estimate from a sample of the population, and calling this the population value. The making an estimate from a subgroup within that sample. The problem here is that the participants in your subsample are double-counted – they are in the estimate for the population value and the estimate for the subsample value.
If you are interested in the number of subjects, objects, or responses which fall into various categories, you could use the chi-square goodness-of-fit test.