Log in / Register
Home arrow Business & Finance arrow Fraud and fraud detection
< Prev   CONTENTS   Next >

Nonsampling Risk

Auditors may draw incorrect conclusions from using sampling techniques where the wrong inferences are not because of the selected samples themselves, but for other reasons not directly connected with the sample contents.

Nonsampling risk is where the auditor may have selected an appropriate sample but arrived at a wrong conclusion. Examples include employing inappropriate audit procedures, failure to recognize errors present, or misinterpreting the results or evidence.

Nonstatistical Sampling Methods

The nonstatistical sampling features built into IDEA are shown in Figure 4.1.

IDEA has easy-to-use random, systematic, and stratified random sampling built in.

Nonstatistical Sampling Feature in IDEA

FIGURE 4.1 Nonstatistical Sampling Feature in IDEA

To use the random sampling method, select Random and enter the number of records you wish to select for sampling. IDEA inputs the start and ending record number based on the records in the data set. IDEA suggests a random seed number (it changes every time you use the random record sampling feature) that you may change if desired to perform the random record sampling extraction. The History file in IDEA records the seed number used. In the example in Figure 4.2, it was decided that approximately 1 percent or 90 records would be examined.

Applying Random Record Sampling in IDEA

FIGURE 4.2 Applying Random Record Sampling in IDEA

IDEA outputs 90 records and creates a SAM_RECNO field that tells you which record number from the original file of 89,979 records it was randomly selected from as displayed in Figure 4.3.

Output of Random Sampling

FIGURE 4.3 Output of Random Sampling

Systematic record sampling extracts a number of records from a database at equal intervals to a separate database. You may input the number of records and IDEA will compute the interval size. In the example in Figure 4.4, we enter 90 records and IDEA calculates the selection interval of 1,010. Conversely, you may enter the selection interval and IDEA will calculate the required number of records.

Applying Systematic Record Sampling

FIGURE 4.4 Applying Systematic Record Sampling

Stratified random sampling is used to extract a random sample with a specified number of records from each of a series of bands. This method requires the database to first be stratified into a series of numeric, character, or date bands. Based on your knowledge of the data set, select the bands' lower and upper limits as displayed in Figure 4.5.

You are then presented with a table displaying the number of records within each band, in which you must decide and enter the number of sample records to be extracted

Selecting the Ranges for Stratified Random Sampling

FIGURE 4.5 Selecting the Ranges for Stratified Random Sampling

at random from each band. In this example, you decide to select more records from higher sales dollar amounts, including 50 records to sample from the 330 records that exceed the maximum upper limit of $100.00 shown in Figure 4.6.

Selecting the Number of Records to Sample in Stratified Random Sampling

FIGURE 4.6 Selecting the Number of Records to Sample in Stratified Random Sampling

Using stratified random sampling eliminates independence of the sample as the auditor makes the judgment to select the sample size from each band or stratum. Care should be taken if employing the results in any statistical application.

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science