a) The number of clients to whom it is appropriate because set up costs and training can be very high;

b) Large populations must exist as statistics is the science of large numbers;

c) Adequate controls must exist and the objective being to test them it is obvious that where no control exist then you cannot apply statistical sampling;

d) The populations being tested must be homogenous in materiality. They must also be homogenous in that the same system and controls must apply to each one of them. In other words they must be subject to the same treatment.

e) Too many variables cannot be tested at once;

f) Items must be separately identifiable, therefore sequential numbering is essential;

g) The error must be defined;

h) Materiality: the auditor must consider the total value of the population and any variances from it;

i) The risk factor: some items have more risk than others;

j) The availability of other evidence. If evidence can be obtained through other means, then statistical sampling may only be a top up.

Statistical sampling - Advantages

a) It is scientific and defensible;

b) It provides a precise mathematical statement about probabilities of being correct;

c) It is efficient as over large samples are not taken;

d) It tends to cause uniform standards among different auditing firms;

e) It can be used by lower grade staff who due to lack of experience may be lacking the necessary judgement needed by the judgement sampling.