AI Fairness and Bias Testing Quiz

Submit

2. True or False: Achieving perfect fairness across all metrics is always feasible in practice.

True

False

3. What is a fairness audit in AI systems?

A systematic evaluation of algorithmic decisions for bias and discrimination

A financial review of AI project budgets

A code review for programming errors

A security assessment of data encryption

Submit

5. Which stakeholder group is critical in defining fairness criteria for AI systems?

Affected communities and domain experts

Only software engineers

Only data scientists

Only legal teams

6. True or False: Transparency in AI systems automatically eliminates bias.

True

False

Submit

8. Which tool is commonly used for detecting and quantifying bias in machine learning models?

IBM AI Fairness 360

TensorFlow Lite

Keras Tuner

PyTorch Lightning

9. What is intersectionality in bias testing?

Examining bias at the intersection of multiple protected characteristics

The point where two datasets overlap

The intersection of accuracy and precision metrics

The overlap between training and test sets

10. True or False: Fairness testing should occur only before model deployment.

True

False

11. What is algorithmic bias in AI systems?

Systematic errors that favor or disadvantage certain groups due to flawed data or design

The deliberate programming of preferences into machine learning models

Random variations in model predictions across different runs

The natural tendency of algorithms to prioritize accuracy over speed

12. Which approach involves testing AI models across multiple demographic subgroups?

Stratified fairness auditing

Ensemble modeling

Cross-validation

Feature scaling

13. What is proxy discrimination in AI systems?

Using correlated variables to indirectly discriminate when direct attributes are unavailable

Deliberately using protected characteristics in model decisions

Refusing to serve certain customer groups

Training separate models for different demographics

14. True or False: Removing sensitive attributes from training data completely eliminates algorithmic bias.

True

False

Submit

16. Which metric evaluates whether false positive and false negative rates are equal across groups?

Equalized odds

Demographic parity

Calibration

Disparate impact ratio

17. What does calibration measure in fairness testing?

The accuracy of predicted probabilities for different demographic groups

The speed at which a model makes predictions

The number of features used in the model

The size of the training dataset

18. True or False: Equalized odds and demographic parity are identical fairness metrics.

True

False

Submit

20. Which of the following is a primary source of bias in training data?

Historical inequities reflected in the data collection process

The number of parameters in the neural network

The computational power used to train the model

The programming language chosen for implementation

SecAI+ Fairness and Bias Testing in AI Systems Quiz

1. Fairness constraints in AI are often ____.

2.

What first name or nickname would you like us to use?

2. True or False: Achieving perfect fairness across all metrics is always feasible in practice.

3. What is a fairness audit in AI systems?

4. Disparate impact occurs when ____.

5. Which stakeholder group is critical in defining fairness criteria for AI systems?

6. True or False: Transparency in AI systems automatically eliminates bias.

7. Counterfactual fairness tests whether model predictions remain consistent when ____.

8. Which tool is commonly used for detecting and quantifying bias in machine learning models?

9. What is intersectionality in bias testing?

10. True or False: Fairness testing should occur only before model deployment.

11. What is algorithmic bias in AI systems?

12. Which approach involves testing AI models across multiple demographic subgroups?

13. What is proxy discrimination in AI systems?

14. True or False: Removing sensitive attributes from training data completely eliminates algorithmic bias.

15. The 80% rule is a legal guideline that checks for ____.

16. Which metric evaluates whether false positive and false negative rates are equal across groups?

17. What does calibration measure in fairness testing?

18. True or False: Equalized odds and demographic parity are identical fairness metrics.

19. Demographic parity ensures that ____.

20. Which of the following is a primary source of bias in training data?