Doug Iceberge, senior-in-charge of the audit of Fisher Industries, has decided to test the following two controls for Fisher's revenue process. 1. All sales invoices are supported by proper documentation, that is, a sales order and a shipping document. 2. All sales invoices are mathematically correct. Iceberge has decided to use a nonstatistical sampling approach based on the following judgments for each control and has judgmentally decided to use a sample size of 50 sales invoice packets. After completing the examination of the 50 sample items, Iceberge noted one deviation for control 1 and two deviations for control 2. What should Iceberge conclude about each control? Nonstatistical Sampling Why do we use non-statistical sampling? - Simplify judgments - Improve consistency Determining the Sample Size - Firm will have established policies - Should be comparable to statistical methods Selecting the Sample Items - Random or systematic selection allowed - Haphazard selection also allowed - Select w/out bias (Is this easy to do?) Nonstatistical Sampling Calculating the Upper Deviation Rate - Cannot mathematically quantify the upper deviation rate and sampling risk - What can you compute? - Sample deviation rate - Expected population deviation rate - Results support level of control risk if sample deviation rate