Fairness Assessment Quiz

What you'll be asked

1. Which of the following is the most common first step when performing a fairness assessment on an ML model?
2. You are evaluating a loan approval model and notice the False Negative Rate is significantly higher for a specific demographic group compared to others. Which fairness metric are you observing?
3. When auditing a model for bias, why is it dangerous to rely solely on 'Demographic Parity' as your only metric?
4. A company wants to ensure their hiring model does not favor any gender. They decide to remove the 'gender' feature from the training dataset. What is the most likely outcome of this action?
5. You are assessing a classifier where the base rate of positive outcomes differs significantly between two groups. You want to ensure that the probability of a positive prediction is the same for both groups, regardless of their actual qualification status. Which metric should you use?

Answers and explanations come after you submit.