Audit Data Quality and Data Analytics AAIA Practice Questions (Page 3)

Questions 11–15

1. 11foundation · easy

   An AI auditor is using a dataset that contains personally identifiable information (PII) for analytics. Which ethical consideration is most important to address when handling this data?

   • ACompleteness
   • BTimeliness
   • CBias
   • DPrivacy
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   During an audit of a recommendation system, the audit team discovers that the 'user_age' field contains several values over 150. The team decides to remediate this issue. Which remediation action is MOST appropriate to maintain data integrity and support the audit trail?

   • ACorrect the ages by subtracting 100 from each value over 150, assuming a data entry error.
   • BDelete the records with invalid ages and note the deletion in the working papers.
   • CReplace the invalid ages with the median age of the dataset and document the change in the audit log.
   • DFlag the records with invalid ages for exclusion from the analysis and document the exclusion criteria.
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   An audit team is using a Python script to analyze a dataset extracted from an AI system's logs. The team needs to ensure that the analysis is reproducible and that the data has not been altered during the process. Which practice is MOST critical to maintain data integrity?

   • ADocumenting the version of Python and all libraries used in the script.
   • BUsing a powerful workstation to run the analysis faster and reduce the time the data is in memory.
   • CCreating a cryptographic hash of the original dataset before starting the analysis and verifying it after the analysis is complete.
   • DStoring the Python script in a shared network folder so other team members can access it.
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   An AI auditor is evaluating a dataset used to train a credit-scoring model. The auditor notices that the dataset contains records with dates in the future and ages that are negative. Which data quality dimension is most directly violated by these issues?

   • AConsistency
   • BValidity
   • CAccuracy
   • DCompleteness
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   An audit of a customer churn prediction model reveals that the 'customer_tenure' field contains negative values, and the 'customer_id' field has some duplicate entries. The audit team needs to prioritize the remediation efforts. Which data quality issue poses the GREATER risk to the model's validity, and why?

   • ABoth issues are equally risky, as they both violate core data quality dimensions.
   • BThe duplicate customer IDs, because they violate the uniqueness dimension and could lead to overcounting in the analysis.
   • CNeither issue is a significant risk, as they can both be easily fixed with data cleansing.
   • DThe negative tenure values, because they violate the validity dimension and could directly skew the model's predictions.
   Reveal Solution Discussion Select an answer first