Which statement is least appropriate?

Several considerations should be applied when deciding when statistical sampling might be appropriate:

a. Only use statistical sampling where it is appropriate. The auditor makes a conscious decision at this stage rather than an instinctive view that it is not normally used.

b. Define and know the population. Where the technique is applied there needs to be a formal process whereby the item that is being considered is fully researched by the auditor.

c. Ensure that every item has an equal chance of selection. Randomness is a main ingredient of statistical sampling as this supports the objective way that the technique should be applied.

d. Ensure that patterns do not affect the randomness. The population should be capable of supporting random sampling in the way it is formed and maintained.

e. Where judgement sampling is used one may still form definite conclusions about the population. The rules on the application of extrapolation mean precise figures can be projected even where there is no scientific basis.

f. Use an error rate that is reasonable. The error rate is built into the statistical tables and is based on assumptions about the population.

g. Stratify the population where this reduces variability. We have touched upon the position where the auditor wishes to follow a certain line of enquiry, and is hindered by the need to assume a neutral stance by the use of statistical sampling.

h. Do not set needlessly high reliability goals. There are accepted standards that reflect the general business environment.

i. Analyse the results carefully. Statistical sampling is a means to an end and results must make sense and fit the audit objective.