During a procurement review, an internal auditor runs an algorithm that flags purchase transactions whose amounts, timing, and approver combinations deviate markedly from the established pattern for that vendor. Which type of data analysis does this technique best represent?
- APredictive analysis, because the model estimates the most likely future value of each purchase transaction from prior vendor behaviour.
- BPrescriptive analysis, because it recommends the specific corrective control the process owner should implement for the vendor.
- CAnomaly detection, because it isolates transactions that diverge significantly from the expected pattern for further examination. Correct
- DText analysis, because it interprets the unstructured narrative descriptions attached to each purchase order line.
Why A is wrong: Predictive analysis forecasts a future outcome or value; it is tempting because both use historical patterns, but flagging deviations from an existing pattern is detection of the unusual, not a forecast.
Why B is wrong: Prescriptive analysis advises what action to take; it is plausible because auditors act on the results, but the technique itself only identifies outliers and does not recommend a control.
Why C is correct: Anomaly detection identifies records that depart from an established norm so they can be investigated, which is exactly what flagging unusual amount, timing and approver combinations does.
Why D is wrong: Text analysis extracts meaning from unstructured language; it is tempting where free-text fields exist, but the technique described works on structured amounts, dates and approvers, not narrative text.