This 3-page paper provides a discussion about sampling procedures used for an accounts receivables of a company. The paper discusses what went wrong with the procedure.
Name of Research Paper File: D0_MTmeadcs.rtf
Unformatted Sample Text from the Research Paper:
sampling in testing controls on Jiffys A/R statements, Mead has opted to use monetary-unit sampling, or MUS, reasoning that MUS would offer a stratified sample, with each account having an
equal chance of being selected for confirmation. Before continuing, we must point out that this assumption is faulty for several reasons. First, MUS
is fine if accounts have similar or the same values. However, the case study points out that the accounts range in size from very small (with recorded balances of less
than $20) to very large - more than the $10,000 that Mead determined the sampling interval should be. Though the study itself doesnt indicate how many total accounts there are
(stating only that Mead selected 60 for review), doing an average with such a wide variance could lead to some troubles. Second, the
main basis behind MUS is that the sample size should be based on the largest amount of the samples indicated - and as mentioned before, the two largest accounts were
automatically scrubbed from sampling, at least, from the start (well see later how Mead ended up "creatively" using these larger accounts to determine mean figures). But the basis of successful
MUS is not only the number of line items in a given population, but also an approximate book value of the largest item - this, as we saw, was ignored.
But Mead did something a little odd later on in the audit process - instead of relying on the small accounts as part
of the sample, the company instead substituted the three largest accounts that had not been originally selected in the sample, but trying to compare the small accounts with the large