Skip to main content

Week 3 Assignment: Bias Case Study Analysis

2 min read (336 words)
#assignment  #bias  #ethics  #week 3 

Week 3 Assignment: Bias Case Study Analysis#

Due: End of Week 3 | Format: Written analysis | Length: 600-800 words

📝 Assignment Overview#

Select a real-world case of algorithmic bias and conduct a deep analysis of what went wrong, why, and what could have been done differently.

📋 Instructions#

Case Study Selection#

Choose one of these cases or propose your own:

  • Amazon Hiring Algorithm — Gender discrimination in recruitment
  • COMPAS Recidivism Tool — Bias in criminal risk assessment
  • Facial Recognition Bias — Misidentification by race/gender
  • Apple Card Algorithm — Gender discrimination in credit decisions
  • Automated Resume Screening — Discrimination by name/background
  • Your choice (get instructor approval)

Analysis (600-800 words)#

Address each section:

1. The System (15%)

  • What problem was the AI system designed to solve?
  • How did it work technically?
  • Who built and deployed it?

2. The Bias (25%)

  • What bias was discovered?
  • How did it harm people? Who was affected?
  • How was the bias detected?

3. Root Causes (25%)

  • Was the bias in the training data, the algorithm, or both?
  • What decisions by the developers led to this outcome?
  • What social/historical factors contributed?

4. Accountability (20%)

  • What were the consequences (legal, reputational, regulatory)?
  • Who was held responsible? Should others have been?
  • What was the outcome/resolution?

5. Lessons & Prevention (15%)

  • What could have been done differently?
  • What process changes or safeguards might prevent similar bias?
  • What does this case teach us about AI ethics?

📌 Rubric#

CriteriaPointsDescription
Case Understanding25%Accurately describes the system, bias, and impact
Critical Analysis40%Thoughtful exploration of causes and implications
Ethical Reasoning20%Connects to course concepts; examines accountability
Writing Quality15%Clear, well-organized, proper citations

💡 Tips for Success#

  • Use multiple sources: original reporting, academic research, company statements
  • Go beyond surface-level description—dig into root causes
  • Connect to course concepts and readings
  • Consider different stakeholder perspectives

📤 Submission#

Submit a PDF document via D2L by 11:59 PM on [Due Date].

Include a reference list with at least 5 sources.

Questions? Post in Q&A or email your instructor.