Artificial Intelligence and Bias

Active learning works well for this topic because AI bias is abstract until students see it in action. When students manipulate datasets or debate real cases, they move from hearing about bias to feeling its impact. This hands-on approach helps them internalize how numbers and decisions interact in ways that create unfair outcomes.

• Grade 11
• Computer Science
• 4 activities
• 30 min–60 min

Learning Objectives

1. Analyze how specific features within a dataset can introduce or perpetuate bias in machine learning models.
2. Evaluate the ethical implications of biased AI decision-making in scenarios such as loan applications or criminal justice.
3. Propose mitigation strategies to reduce bias in an AI model, considering trade-offs between fairness and accuracy.
4. Critique existing AI applications for potential biases and their societal impact.
5. Explain the concept of algorithmic fairness and its challenges in diverse social contexts.

Case Study Analysis

⏱ 45 min·Small Groups

Case Study Analysis: Real-World AI Bias

Provide articles on cases like COMPAS recidivism tool or Amazon hiring AI. In small groups, students identify biased data sources, predict outcomes, and propose fixes. Groups present findings to class for peer feedback.

Prepare & details

Who is responsible when an autonomous system makes an unethical decision?

Facilitation Tip: During Case Study Analysis, assign roles like 'data scientist' or 'ethicist' to push students beyond surface observations.

Setup: Groups at tables with case materials

Materials: Case study packet (3-5 pages), Analysis framework worksheet, Presentation template

• Analyze
• Evaluate
• Create
• Decision-Making
• Self-Management

Socratic Seminar

⏱ 30 min·Pairs

Dataset Audit: Spot the Bias

Give pairs anonymized datasets from hiring or lending scenarios. They categorize features, calculate representation imbalances, and graph disparities using spreadsheets. Discuss how imbalances affect model training.

Prepare & details

How can we detect and mitigate bias in algorithmic decision-making?

Facilitation Tip: For Dataset Audit, require students to document each bias they find with a specific example from the dataset.

Setup: Chairs arranged in two concentric circles

Materials: Discussion question/prompt (projected), Observation rubric for outer circle

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Socratic Seminar

⏱ 50 min·Whole Class

Simulation Debate: Ethical Decisions

Whole class divides into roles: developers, users, regulators. Simulate an AI car accident scenario with biased training data. Debate responsibility and mitigation, voting on solutions.

Prepare & details

What does it mean for a machine to be 'fair' in a social context?

Facilitation Tip: In Simulation Debate, provide a clear rubric for ethical frameworks so students can ground their arguments in evidence.

Setup: Chairs arranged in two concentric circles

Materials: Discussion question/prompt (projected), Observation rubric for outer circle

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Socratic Seminar

⏱ 60 min·Individual

Bias Mitigation Workshop: Model Tweaks

Individuals tweak a simple pre-built ML model (using Google Colab) by resampling data or adding fairness constraints. Test on holdout sets and compare accuracy vs. fairness metrics.

Prepare & details

Who is responsible when an autonomous system makes an unethical decision?

Facilitation Tip: During Bias Mitigation Workshop, circulate with a checklist to ensure teams test at least two different strategies.

Setup: Chairs arranged in two concentric circles

Materials: Discussion question/prompt (projected), Observation rubric for outer circle

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Teaching This Topic

Experienced teachers approach this topic by balancing technical details with ethical questions. Avoid diving too deep into machine learning math, which can overshadow the human impact of bias. Instead, use analogies like 'training data as a mirror' to help students grasp how society’s flaws become AI’s flaws. Research shows that students retain concepts better when they connect them to lived experiences, so encourage personal reflections on fairness.

What to Expect

Successful learning looks like students confidently identifying bias in datasets and explaining its origins. They should also evaluate responsibility for AI decisions and propose concrete mitigation strategies. Most importantly, they should connect these concepts to fairness in technology and society.

Watch Out for These Misconceptions

Common Misconception

During Simulation Debate, watch for students who assume AI models are neutral due to their mathematical foundations.

What to Teach Instead

Use the debate format to redirect them: ask teams to present evidence from their case studies showing how training data shapes model behavior, then have peers challenge each claim with dataset examples.

Common Misconception

During Dataset Audit, watch for students who dismiss subtle biases as unimportant.

What to Teach Instead

Guide them to trace correlations step-by-step, using the activity’s audit sheet to mark how small imbalances grow into larger disparities over time in predictive models.

Common Misconception

During Bias Mitigation Workshop, watch for students who believe bias cannot be fixed once a model is trained.

What to Teach Instead

Have them experiment with techniques like data reweighting during the workshop, then compare results to show how intervention changes model behavior, even post-training.

Assessment Ideas

Discussion Prompt

After Case Study Analysis, present the job recommendation scenario. Ask students to use evidence from their case studies to justify who bears responsibility and what steps developers should take, then facilitate a whole-class vote on the most compelling arguments.

Quick Check

During Dataset Audit, collect students’ completed audit sheets and review them for at least two correctly identified biases and explanations of how those biases could affect predictions. Use this to adjust the workshop focus.

Exit Ticket

After Bias Mitigation Workshop, ask students to write one example of bias entry and one mitigation strategy they tested, along with a definition of 'algorithmic fairness' in their own words, before they leave the classroom.

Extensions & Scaffolding

• Challenge advanced students to design a new mitigation technique and test it on a real dataset.
• Scaffolding for struggling students: Provide a partially completed bias audit worksheet with examples to guide their analysis.
• Deeper exploration: Invite a guest speaker from tech ethics or have students research a well-known case study like COMPAS recidivism predictions.

Key Vocabulary

Algorithmic Bias	Systematic and repeatable errors in a computer system that create unfair outcomes, such as privileging one arbitrary group of users over others.
Training Data	The dataset used to teach a machine learning model to recognize patterns and make predictions. Biases present in this data can be learned by the model.
Fairness Metrics	Quantitative measures used to assess whether an algorithm's outcomes are equitable across different demographic groups, such as demographic parity or equalized odds.
Data Augmentation	Techniques used to increase the size and diversity of a training dataset, often by creating modified versions of existing data, to help reduce bias.
Algorithmic Accountability	The principle that developers and deployers of AI systems are responsible for the outcomes and impacts of their algorithms, especially in cases of harm or discrimination.