In watsonx.governance fairness monitoring, what does the disparate impact ratio measure?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Think of it this way: in real-world AI governance, the ratio of the favorable outcome rate for the monitored group divided by the favorable outcome rate for the reference group is exactly what teams reach for when they need to handle this scenario. The disparate impact ratio divides the monitored group's favorable outcome rate by the reference group's favorable outcome rate. On the exam, remember that this falls squarely under the 4.0 Configure Evaluation and Monitoring domain.
Full explanation below image
Full Explanation
The disparate impact ratio divides the monitored group's favorable outcome rate by the reference group's favorable outcome rate. A value below 0.8 indicates potential adverse impact under the four-fifths rule. It differs from statistical parity difference, which expresses the same comparison as an absolute difference rather than a ratio. The correct answer, "The ratio of the favorable outcome rate for the monitored group divided by the favorable outcome rate for the reference group", directly addresses the scenario described because it aligns with the specific governance requirement in question. The incorrect options ("The difference between the true positive rates of the monitored group and the reference group", "The absolute difference in selection rates between demographic groups in the dataset", "The ratio of false positive rates between the monitored group and the reference group") may seem plausible but do not satisfy the core requirement. Understanding the distinction between these concepts is critical for IBM watsonx.governance implementations and is frequently tested in the 4.0 Configure Evaluation and Monitoring section of the certification exam.