ISA 530 states that for tests of controls, no explicit projection of errors is necessary because the sample deviation rate is also the projection deviation rate for the population as whole. So for example, if in sample of 75, four errors are discovered, the projected deviation rate is 4/75, ie 5%.