A is corrent because sampling for variables is used in substantive testing to estimate the dollar amount of a population or the amount of error in a population.
B is incorrect because sampling for attributes is used in tests of controls to estimate occurrence rates in a population.
C is incorrect because numerical sampling has no established meaning for audit sampling.
D is incorrect because discovery sampling is a type of attribute sampling used in tests of controls when very low occurrence rates are expected and the auditor wants a chance to observe one occurrence.
| 
