Internal auditors make judgments about audit issues or fulfill their audit objectives through detailed reviews of what is called audit evidence. In particular, given today's IT systems, internal auditors need some way to review this large mass of computerized data. Internal auditors need a consistent approach to sample a portion of items from a large data population and then to draw audit conclusions based on that limited sample. The challenge of internal audit sampling here is extracting a sample of items that will be representative of the entire population. Audit sampling has two main branches: statistical and non-statistical. Statistical sampling is a mathematical-based method for selecting representative items that reflect the characteristics of an entire population. Non-statistical sampling, also called judgmental sampling, is not supported by mathematical theory and does not allow internal auditors to express statistically correct opinions across the population.
10.2 AUDIT ASSESSMENT AND EVALUATION TECHNIQUES
When planning any inspection that includes examining a large number of transactions or other evidence, internal auditors should always ask the question: Should I use audit sampling? The correct answer here is often not only yes or no but may be complicated by factors such as the amount or nature of the items to be. sampling, lack of expertise or technical availability of computer software to perform sampling, fear of focused mathematical sampling, and potential non-acceptance of sampling results by management. To develop audit conclusions based on such data, internal auditors need a process by which they must: A. Understand the total population of items of concern and develop a formal sampling plan regarding the population of items. B. Describe a sample of the population based on the sample selection plan. C. Evaluating sample items against audit objectives. D. Develops conclusions for the entire population based on the results of the audit sample. Formal Audit Sampling is a powerful tool, and with some education and practice, internal auditors can start using it easily and effectively. The following reasons encourage the use of audit sampling and sampling statistics in particular include: Conclusions can be drawn about the entire data population Objective and defensible sample results Less sampling may be required through the use of audit sampling Statistical sampling can provide more accuracy greater than the 100% test Multiple site audit coverage is often more convenient Sampling procedures can be easily implemented.
10.3 INTERNAL AUDIT JUDGMENTAL SAMPLING
Although the author encourages a statistical audit sampling approach, nonstatistical sampling is a very appropriate internal audit procedure in many situations. Judgmental sampling requires internal auditors to select samples that are representative of the items in the population or transactions for audit review. For internal auditors, methods for selecting judgmental sampling may take many forms, including: Fixed percentage selection Defined attribute selection Many options - values Designated area selection Other selected attribute options. Although useful data can be obtained from judgmental sampling, the results can be misleading or inaccurate regarding the entire population or me. Internal auditors can assess the accuracy of financial charges for the greatest 10% of several accounts under the assumption that this is the most significant. When planning a review based on a judgmental sample, the internal auditor must make three judgmental sampling decisions. First, the internal auditor must develop a selection method and decide what types of items to examine. The sample size is an examination judgment decision. Second, the sample size must be reasonable in comparison with the entire population. A sample that is too small will not represent the entire population, whereas a sample that is too large may be too time-consuming or expensive to evaluate. The third decision is how to interpret and report the audit results from limited judgmental sampling. The whole concept behind internal audit judgmental sampling is that item selection is based on internal auditors' judgment. An internal auditor may select as many or as many sample items that appear appropriate in his or her professional judgment.
10.4 STATISTICAL AUDIT SAMPLING: AN INTRODUCTION
Statistical sampling is a powerful tool that allows internal auditors to project the results of an audit sample over the entire population with a strong degree of accuracy and confidence. Based on probability rules, statistical sampling requires the use of established mathematical selection techniques with results that can be projected across the population in a way that courts, government regulators, and others will accept. The following is presented to help internal auditors use statistical sampling more effectively: Statistical Sampling Concepts Some of the sampling concepts are fairly straightforward, and a general understanding is important. We start with some important statistical sampling terms. First, the word population refers to the number of items subject to an audit, and random sampling is the process of selecting a sample whereby each unit in the population has an equal probability of selection. The random sample of 10 must represent one characteristic of the entire population. However, the characteristics of one random sample drawn by internal auditors may differ from samples from the same population drawn by others. Since several samples can produce different results, it is important to understand the term statistical sampling for measures of central tendency. In sampling examination, the mean value of expression is used to describe the characteristics of the sample data set. The most common statistical measures for viewing data are the mean, median, mode, range of data values, variance, standard deviations, and slope of the data. The mean is the simple average of the value of the items in a population Median is the mean number of items in the population Mode is the number or value that occurs most frequently in the population Range is the difference between the largest and the smallest values in a population. Variance is a measure of the spread of the distribution, and is calculated as the mean of the squared deviation of Most professionals should find each of the steps discussed so far to be fairly easy to understand, even those without much background in statistics The standard deviation tells the auditor how much many variations in value exist
10.5 DEVELOPING A STATISTICAL SAMPLING PLAN
An internal auditor must understand the nature of the data to be reviewed when developing a strategy selection of this sample or audit plan, including: The population to be sampled must be clearly defined. The population is the total number of units from which a sample can be drawn, including scope or the nature of the item to be reviewed The population should be divided or grouped into groups if any large variation exists among population items. Internal audit may have to consider the stratification of the sample by placing small sets of high- value items in one population and equilibrium in separate populations. Separate samples can be taken from each. A. Every item in a population must have an equal chance of being selected B. Internal audit must always disclose facts when reporting results. C. There should be no bias in making a sample selection from the population. D. In conclusion the sample results of the reviewed local items may not be representative of the remote warehouse. There are four general methods for selecting an audit sample: random number, interval, stratification, and cluster selection. Modern internal auditors must have a general understanding of most techniques appropriate for the specific audit situation. Random Number Audit Sample Selection The items here must be randomly selected, with each one in the population having its own equal chance of being selected as part of the sample. Although the main public accounting firm have developed a minimum theoretical sample size for their testing (often 60 or even 30), internal auditors generally have to use no minimum sample size in the audit. Interval Selection Audit Sample Selection This technique is especially useful for sampling monetary units, in which an auditor internal will develop the sample by selecting each nth item in the population. Need that there is fairly homogeneous in the population, in terms of types of goods, and there is no internal bias population settings that would result in a sample that was not statistically representative. Stratified Selection Audit Sample Selection Multilevel selection is an extension of the random or interval selection technique, as neither can applied smaller population strata. In some cases, one of the strata can be examined 100% while others will be subject to random selection. The justification for this stratification is possible that one stratum has very different characteristics, and that internal audit may be desirable to evaluate these subgroups more individually and precisely. Cluster Selection Audit Sample Selection Samples are made by systematically selecting subgroups or groups of the total population.
10.6 AUDIT SAMPLING APPROACH
The three most common approaches are attribute sampling, variable sampling, and sampling discovery. The sampling attribute is used to measure the rate at which various conditions occur or attributes in other words, to assess internal control. Attributes or characteristics can be applied on any physical item, financial record, internal procedure, or operational activity. Sampling The variable relates to a specific population size, such as account balances or tests in sample items individual. Variable sampling is concerned with the absolute number as opposed to the number or level of certain types of errors. Discovery sampling is used when the internal auditor wants to draws a sample of large amounts of data without statistical processing of the variables and attributes of the sample. 1. Attributes Sampling Procedures Attribute sampling is the process of drawing samples to estimate proportions some of the characteristics or attributes of interest in a population. The starting point in the attribute sampling is estimating the expected error rate it is, how much errors that internal audit and management can tolerate? Depends on the sample item and corporate culture, this expected error rate may be as little as 0.01% or as large 5% or even more. 2. Performing Attributes Sampling Test Attribute sampling is useful when internal auditors are faced with numbers a rather large item to be examined and want to test whether a particular control is functioning or not working. The auditor must first determine what to evaluate or the specific nature of the compliance tests to be performed, the nature of the sampling unit, and the population characteristics.
10.7 AUDIT SAMPLE SAMPLE ATTRIBUTES
Gnossis has about 4,000 employees, and internal audit has decided to do Attribute tests to assess internal controls that include human resource records. Source function Human power Gnossis uses two IT systems to record one employee for calculations payments and one to benefit and manage a desktop based spreadsheet system for things such as an employee health insurance declaration. Through a review of the human resource recording process, Internal audit found about 30 different record-keeping control issues, ranging from large to large like whether wages are deducted properly for tax purposes to what small things. Monthly deductions for paying employee credit union contributions are correct. Internal audit has combines all 30 of these recording problems as one attribute, as a single yes or no test. The problem here is that some minor issues will force internal audit to conclude that the internal controls do not work even though no problem is found over the main problem. This will often be difficult to communicate to management. The audit strategy was to test the Gnossis human resource records for separate attributes. Although internal audit can be tested separately for all 30 attributes, a better approach is to decide which are the most significant and to test only for those separate attributes. Assume that internal audit has decided to test human resource records for the following five attributes: A. The payment value and status on the automatic system must be the same as the one in the manual file. B. Authorization for deduction must be signed and dated by the employee. C. The preemployment background check should have been completed. D. If there is no deduction for life insurance, the employee's signed waiver must be recorded. E. Salary increases are in accordance with the guidelines and are well authorized
10.8 Attributes Sampling Advantages and Limitations
Sample attribute advantages: 1. When there is a need to review a large number of items, attribute the sampling procedure can provide a statistically accurate assessment of the features or attributes of a control 2. A very useful tool for reviewing control procedures in various operational areas. Some internal auditors feel this technique has several obstacles in its use, including: The attribute sampling calculation is complex. This chapter has just introduced some very basic attribute sampling concepts. Auditor internals need to have a good understanding of the process or could be in danger of interpreting it results incorrectly. The exact definition of attributes can be difficult. The selection of the attributes to be tested is based on the auditor's assessment or management's request. However, an auditor may overlook one or more other important attributes when analyze data.
10.9 Monetary Unit Sampling
Attribute sampling measures the extent of several conditions, and variable sampling estimates the value of an account. Monetary unit sampling is a technique for determining whether an account is financial is fairly presented, and is a good method of estimating total accounts overestimated. The concept is that every dollar or currency unit is denominated in an account is treated as part of the population and each has the opportunity to selected. Selecting Monetary Unit Samples An internal auditor can select a sample using a spreadsheet program or via manual calculations using a calculator. Its purpose is to determine monetary distance based on the calculated sample size. There are two key points and restrictions on the unit monetary sampling. First, monetary unit sampling is useful only for testing the existence of a statement which is excessive. Second, the selection method described does not handle credit amounts correctly. The best solution is to withdraw all recorded and treated credit balances as distinct populations to evaluate. Carry out Monetary Unit Sampling Testing The dollar number evaluated in a population is the sample size of auditors. Same with attribute sampling, testing monetary unit sampling requires 4 things about that account will be sampled: 1. The maximum percentage error of the recorded population value that the auditor can accept. 2. The expected level of confidence. 3. The expected error rate for sampling error. 4. The total carrying amount of the accounts being evaluated. The monetary unit sample size is then used to calculate the monetary distance with divide the carrying book value of the accounts by the sample size to determine each n interval dollar. This interval specifies the selection limit for the larger item, and for all items which is equal or greater to this interval will be selected. Evaluating the Results of Monetary Unit Samples Monetary unit sampling is a popular approach to evaluating balances account to determine if the value is overestimated. Because every dollar is inside Each item an account will be subject to sample selection, the items are presented too height will be found during the sampling process. The basic idea is to document the carrying amount and audited amount for each selected item and then calculating the respective percentage error. Advantages and Disadvantages of Monetary Sampling Units The most important advantage of monetary unit sampling is that it focuses on units of greater value based on random selection. Because monetary unit sampling selects a sample item online proportionally, there is less risk of detecting a material error due to all high-value units are subject to selection according to their respective sizes. The main drawback of sampling is that the procedure does not adequately test reports the finances presented are too low.
10.10 Other Audit Sampling Techniques
Internal auditors must develop a high-level general CBOK understanding of the technique this, but building expertise takes some extra work. Sampling, however, is a large area; other less complex methods may be used in certain circumstances. Part The next section briefly describes some of these internal audit sampling methods other. Multistage Sampling Multistage sampling involves sampling at several levels. A random sample is chosen first for some unit group, and then another random sample is drawn from within that population first selected. Replicated Sampling Replicated sampling is a variation of multistage sampling that requires a withdrawal from one overall random sample of size X, consisting of a random subsample of Y separate from size X / Y. Bayesian Sampling A technique rarely used or mentioned in the audit sample literature but has Big potential is Bayesian sampling. This procedure is named after a mathematician English Thomas Bayes and is based on an improved probability of sample size and the like, using what is called a subjective probability obtained from the test results
10.11 Making Efficient and Effective Use of Audit Sampling
Internal auditors will or will not decide to test transactions when conduct audits. Internal auditors decide, on the basis of overall comparisons and other audit procedures, that test transactions are not required or the amounts involved are not enough material to warrant testing. If a test transaction is carried out through an audit sample it indicates that it is operating acceptable, no further work required. Where errors are found, Internal auditors must consider one of the following methods in order to achieve this audit conclusion: Splitting errors. Through a review of the types of errors and their causes, internal the auditor will be able to separate the number of errors. Only reports the items that were tested. When the internal auditor faces a significant error, will need to report test results to operating personnel. Perform 100% audits. Although internal auditors are not expected to do detailed testing of all transactions, sometimes it is necessary to extend inspection when it finds significant errors. Projecting sample results. If item selection is made on a random basis, the results can be evaluated statistically. Error numbers and dollar amounts can be predicted for determines the error distance at the given confidence level