引導式測驗 · 中等 等級
AI審計風險評估與規劃測驗
測試日記帳分錄評分和外部訊號如何回饋審計計劃、為什麼風險評分不是結果以及何時仍需要實質程序。
問題 1 的 8
A journal entry model gives an entry its highest risk score. What does that score mean?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- A journal entry model gives an entry its highest risk score. What does that score mean?
- The entry is a confirmed misstatement
- The entry must be reversed by the client
- The entry is a reason to investigate, not a finding of misstatement
- The entry is excluded from sampling
- An account scores low on every analytic but is a significant account. What does AS 2301 still require?
- Substantive procedures for each relevant assertion
- Nothing further if controls are effective
- Only inquiry of management
- Removing it from the audit scope
- Why is mapping a client's chart of accounts to a standard taxonomy important?
- It is required by Benford's law
- It lets models and peer comparisons treat similar accounts consistently
- It replaces the need to test data completeness
- It encrypts the ledger for privacy
- Receivables grew much faster than revenue compared with peers' public filings. Which assertions does the guide link to that signal?
- Presentation of equity
- Completeness of payroll
- Occurrence of lease payments
- Existence and valuation of receivables
- Why do scoring systems often combine rules with unsupervised anomaly detection?
- Either approach alone tends to miss risks or flag too much
- Rules are banned in PCAOB audits
- Unsupervised models always need labeled fraud cases
- Combining them removes the need for thresholds
- What should thresholds for flagging entries connect to so teams are not flooded with alerts?
- The number of users in the ERP
- The client's stock price
- Performance materiality
- The number of staff on the engagement
- Which check confirms that the journal entry data used for scoring is complete?
- Counting the number of distinct users
- Rolling entries forward from the opening trial balance to the closing trial balance
- Running Benford's law on descriptions
- Asking the CFO to sign the extract
- A new product line changes how revenue should be recognized. Why might an analytics model miss this risk?
- Models cannot read numbers
- Revenue is never analyzed by models
- Models only work on prior-year data from other clients
- The model lacks business context unless people supply it