The correct answer is: It may be necessary to increase the sample sizes for tests of controls; It may be possible to find alternative internal controls to evaluate and test.

Where control risk is assessed as high, sample sizes for tests of controls will increase, or alternatively it may be possible to find alternative internal controls to evaluate and test.

It will not always be necessary to take a wholly substantive approach, particularly if alternative controls can be found and substantive testing will only need to be increased if tests of controls prove that control risk is high. It is unlikely that sample sizes for tests of controls will be reduced.