Lessons · Accounting · testing some of it, and what that costs you
Testing some of it, and what that costs you
Auditors test samples because testing everything is impossible. The cost is sampling risk: the chance that the sample is not representative and the conclusion drawn from it is wrong about the population.
Hone is a place to practise a career, one idea a day. This is one of its lessons, written out in full and free to read without an account.
In beta. This lesson was written for Hone and has not yet been checked by a qualified accountant. Practice material, not professional advice. What that means.
What it is for
It is where reasonable assurance actually comes from. Sampling risk is not a flaw to be apologised for; it is a quantity to be managed, by taking larger samples where the risk is higher and by choosing items in a way that does not systematically miss the ones that matter.
How to think about it
Set the sample by the risk and by the working threshold, not by habit. Then decide how items are chosen, so that unusual items are not systematically excluded.
Worked example
Higher risk means a larger sampleMore work where being wrong is likelier.
A lower working threshold means a larger sampleSmaller errors matter, so you need to see more.
Selection must not exclude the odd itemsPicking only round numbers, or only large ones, misses what is interesting.
Sampling risk remains, and is statedIt is managed, not eliminated.
Your turn
Name the risk that comes from testing only part of a population.
The chance a sample is not representative is risk
Solve one, graded on the server
The trap
Treating a sample that produced no errors as proof there are none. It is evidence that the error rate is probably low, which is a different and much weaker statement, and it is the whole reason a margin is kept below materiality.