decision 9

Reminder - I do not allow direct quotes in student's work in this class.  You need to paraphrase and cite your outside sources.  "Outside sources" do not include the text and so you don't need to cite it.

All these questions are going to force you to get into the underlying math of attribute sampling, MUS, and VS.  Thus, they are going to be challenging.  My goal is to have a discussion that will build some intuitions about what the math really does and how it effects how these sampling techniques are applied. 

1)           Provide an intuitive description of MUS in your own words.  Specifically, include how it relates to attribute sampling and yet provides evidence about monetary account balances. 

Hint:  MUS and attribute sampling are based on exactly the same math and distributional assumptions.  Both assume that items in the population can only take two values - right or wrong.  However, while attribute sampling tests controls that can either work or not, MUS tests items in an account balance that can be in error by varying amounts.  I want you to think about what techniques and assumptions MUS uses to adapt attribute sampling for an item that can vary continuously. 

2)           Why does the text emphasize that MUS is biased towards overstatement errors and not understatement errors?  Does variables sampling (VS) contain this bias?  Be sure to explain your answers in your own words.

Hint:  Understanding MUS's bias is important because auditors are sometimes more interest in understatement errors (e.g., in liability accounts) than overstatement errors.  To understand MUS's bias, think about how a sample is selected.  How MUS selects samples is an artifact of issues that should come out in the first question about how it modifies attribute sampling to test balances. 

3)           Describe in your own words the strengths and weaknesses of variable sampling (VS) versus MUS and the conditions where the auditor might choose one over the other.

Hint:  One issue that should come out here flows from question 2).  Also, consider issues like what modifications need to be made to attribute sampling to make MUS work and how VS is designed to tests amounts that can vary continuously.  I will be honest that one difference the text mentions with which I disagree is that MUS is cheaper and easier to apply than VS.  But, I will expand on that in the discussion. 

a) Strengths of MUS versus VS, when MUS is more appropriate and why.

b) Strengths of VS versus MUS, when VS is more appropriate and why.