Data Analytics
Data analytics is a substantially different approach than exploratory analysis.
With exploratory analysis, the purpose is to gain a thorough understanding of the population’s main characteristics. Information from exploratory analysis provides necessary information for developing a sampling plan. With data analytics, the intent is much broader. It can be used, for instance, to discover trends in a population, to detect anomalous activities and identify high-risk items, or to make inferences and predictions about a population. Leveraging artificial intelligence and other technological innovations, many software applications have been developed to conduct increasingly sophisticated analyses of population data in an audit context.
One main advantage of modern data analytics applications is that they can allow auditors to analyze 100% of the units in a population, thus making sampling unnecessary. However, this is only possible when a number of requirements are met. Ultimately, data analytics is dependent on the availability of data that is
- reliably collected,
- provided in a format that enables analysis by the auditors’ software, and
- of sufficient comprehensiveness that it allows for making audit observations and conclusions.
In situations where administrative data is incomplete, poorly organized, or mostly available in non-electronic formats, a data analytical approach will be of little value and traditional sampling approaches will remain useful tools for auditors.
However, depending on the circumstances, it may be possible to combine data analytics and sampling methods. For example, it may be possible to use data analytics to do preliminary population analysis and identify groups of outliers or high-risk cases and then to use a sampling approach to examine in detail a number of these cases. (For a more extensive discussion of data analytics in performance audits, see European Court of Auditors, 2020.)
LinkedIn
Twitter