Other Factors to Consider When Selecting a Sampling Approach

The level of homogeneity of the population

Is the population relatively homogenous? If it is, a generalizable sampling approach using a simple random sampling strategy, or a more sophisticated method may be adequate. Otherwise, a purposeful sampling might be more appropriate. When auditors analyze the population, they will collect information that is key to making the right decision about strategy.

The availability of corroborating evidence obtained from other sources

When determining the sampling approach, like any other evidence collection strategy, auditors consider the global context of the audit approach. Is the evidence resulting from the sample indispensable to conclude on the audit objective? Does it complement other sources of evidence, or does it have to be interpreted in conjunction with other qualitative (testimonial or documentary) evidence to provide a more complete and convincing picture of the program’s performance? The answers to these questions require considerable professional judgment that must take into account the evidence collection strategy as a whole.

The availability of statistical expertise and software

The ability to properly develop complex sampling plans is helped by the availability of statistical expertise, on staff or acquired through contracts. Practice guides, such as this one, or other resources (including those in the Bibliography section of this guide) are of course good tools to guide auditors, but they also have limits. In some instances, auditors may be limited in their capacity to implement more sophisticated approaches, either because of resource constraints or because they do not have an appropriate statistical software (or do not know how to use their software properly).

A specialized software application is a practical tool to assess key population characteristics and calculate sample sizes and other sampling parameters. Some information on this type of software is in Appendix 1.

The time and resources available to complete the audit work

All the above factors have a time or a resource component. Does the audit team have time to use the optimal sampling approach? Can a more modest approach achieve the same goal? Can the audit team afford hiring the expertise needed or purchasing a licence for the required software? Overall, does the audit team have enough resources to examine the number of sampling units needed to meet the requirements to conclude on the audit objective with sufficient and appropriate evidence? Ultimately, the decision may come down to a cost–benefit analysis.²

To guide auditors when they navigate these factors, Table 3 summarizes the advantages and disadvantages of generalizable and purposeful sampling approaches.

Table 3 – Advantages and Disadvantages of Generalizable and Purposeful Sampling Approaches

Source: Modified from The Internal Auditor Professional Services Limited (n.d.), Sampling for Effective Internal Auditing, and from J. L. Colbert (1991), Statistical or Non-statistical Sampling: Which Approach Is Best?

² Auditors can rely on a sample already extracted by other auditors, assuming that it was done for a sufficiently compatible audit on the same population. The decision to rely on such a sample should be supported by an analysis of the sampling plan. (See the section on Documenting the Plan and Its Results for a list of items that should be documented.)

• PreviousDefining the Objective of the Sample
• NextStep 3 – Preparing a Sampling Plan