When monetary-unit sampling fits a receivables population
Trade receivables in mid-market Taiwanese manufacturers often show a familiar shape: a handful of large buyers and a long residual of small invoices. Monetary-unit sampling (MUS) pulls attention toward the larger book balances, which is usually where existence and valuation risk concentrate.
Classical attribute sampling still has a place when you care about the rate of invoice-level errors rather than dollar coverage — for example, when testing cut-off stamps or authorization signatures. Mixing the two without documenting why creates confusion at review time.
Before choosing MUS, confirm the population is free of credit balances that would distort selection, and decide how zero-balance accounts will be handled. Write those decisions into the sampling memorandum; reviewers in Taichung and Taipei firms routinely ask for them.
If prior-year testing found systematic pricing differences in a customer segment, consider stratifying that segment out of the MUS population and testing it separately. Stratification is not decoration — it is how you keep sample sizes honest.