Algorithmic Bias Detectives: Exposing AI's Hidden Biases
Inquiry Framework
Question Framework
Driving Question
The overarching question that guides the entire project.How can we, as Algorithmic Bias Detectives, design and advocate for AI applications that promote fairness and equity in our increasingly AI-driven world?Essential Questions
Supporting questions that break down major concepts.- How can we ensure AI systems are fair and unbiased?
- What factors can contribute to bias in AI algorithms and datasets?
- How does algorithmic bias impact individuals and society?
- What are the ethical considerations surrounding the use of AI in everyday life?
- How can we address and mitigate algorithmic bias in AI applications?
Standards & Learning Goals
Learning Goals
By the end of this project, students will be able to:- Understand sources of bias in data.
- Evaluate the impact of algorithmic bias on society.
- Develop strategies to mitigate bias in AI applications.
- Understand how AI is used in everyday applications.
- Evaluate ethical considerations for AI development and use.
- Communicate findings and recommendations effectively.
Entry Events
Events that will be used to introduce the project to students'Future Headlines' Brainstorm
Students brainstorm and create 'future headlines' predicting the impact of AI on various aspects of society (jobs, healthcare, etc.). This activity encourages them to consider ethical implications and potential biases embedded in these future scenarios.'The Case of the Unfair Algorithm'
Introduce a fictional 'case study' where an algorithm makes biased decisions. Students work in teams as 'AI detectives,' analyzing data and interviewing 'witnesses' to uncover the source of the bias and propose solutions for fair outcomes.Portfolio Activities
Portfolio Activities
These activities progressively build towards your learning goals, with each submission contributing to the student's final portfolio.'Bias Baseline' Brainstorm
Students brainstorm all the different types of biases they have heard of. Then the teacher leads a discussion about how those biases could appear in data.Steps
Here is some basic scaffolding to help students complete the activity.Final Product
What students will submit as the final product of the activityA list of potential biases in data, categorized by type.Alignment
How this activity aligns with the learning objectives & standardsLearning Goal: Understand sources of bias in data.'Algorithmic Autopsy'
Students dissect a pre-built AI application (e.g., a simple image classifier) to identify potential sources of bias in its training data or algorithms.Steps
Here is some basic scaffolding to help students complete the activity.Final Product
What students will submit as the final product of the activityA report outlining potential sources of bias in the AI application and its dataset.Alignment
How this activity aligns with the learning objectives & standardsLearning Goal: Evaluate the impact of algorithmic bias on society.'Fairness Fix-It'
Students propose and prototype solutions to mitigate the biases identified in the previous activity, such as re-weighting data or modifying the algorithm.Steps
Here is some basic scaffolding to help students complete the activity.Final Product
What students will submit as the final product of the activityA revised AI application with implemented bias mitigation strategies and a report on the impact of these strategies.Alignment
How this activity aligns with the learning objectives & standardsLearning Goal: Develop strategies to mitigate bias in AI applications.'AI in My Life' Scavenger Hunt
Students conduct a scavenger hunt to identify AI applications in their daily lives (phones, streaming services, etc.) and document their findings.Steps
Here is some basic scaffolding to help students complete the activity.Final Product
What students will submit as the final product of the activityA presentation (poster, slideshow, or video) showcasing their AI scavenger hunt findings, with a focus on ethical implications.Alignment
How this activity aligns with the learning objectives & standardsLearning Goal: Understand how AI is used in everyday applications.'Ethical AI Dilemmas' Debate Prep
Students prepare for a class debate by researching ethical dilemmas related to AI, such as privacy concerns, job automation, and biased decision-making.Steps
Here is some basic scaffolding to help students complete the activity.Final Product
What students will submit as the final product of the activityDebate preparation materials, including research notes, arguments, and counterarguments.Alignment
How this activity aligns with the learning objectives & standardsLearning Goal: Evaluate ethical considerations for AI development and use.'Bias Busting' Public Service Announcement
Students create a public service announcement (PSA) to educate others about algorithmic bias and its impact on society.Steps
Here is some basic scaffolding to help students complete the activity.Final Product
What students will submit as the final product of the activityA public service announcement (PSA) educating others about algorithmic bias and its impact on society.Alignment
How this activity aligns with the learning objectives & standardsLearning Goal: Communicate findings and recommendations effectively.Rubric & Reflection
Portfolio Rubric
Grading criteria for assessing the overall project portfolioAlgorithmic Bias Detectives Portfolio Rubric
Understanding Bias
Demonstrates comprehension of the nature, sources, and impact of bias in AI systems and data.Identification of Biases
Accurately identifies and categorizes various types of biases in data and algorithms.
Exemplary
4 PointsIdentifies and provides insightful examples of a wide range of biases, demonstrating a nuanced understanding of their origins and manifestations.
Proficient
3 PointsIdentifies and categorizes common types of biases with clear examples, showing a good understanding of the concept.
Developing
2 PointsIdentifies some types of biases but struggles to provide clear examples or accurate categorizations.
Beginning
1 PointsShows minimal awareness of different types of biases and their relevance to data and algorithms.
Impact Analysis
Analyzes the impact of algorithmic bias on individuals, groups, and society.
Exemplary
4 PointsProvides a comprehensive analysis of the far-reaching consequences of algorithmic bias, including ethical, social, and economic impacts.
Proficient
3 PointsAnalyzes the impact of algorithmic bias on various groups and provides relevant examples.
Developing
2 PointsDescribes the impact of algorithmic bias in basic terms or provides superficial analysis.
Beginning
1 PointsShows almost no understanding of the impact of algorithmic bias on individuals or society.
Mitigation Strategies
Proposes and implements effective strategies to mitigate algorithmic bias.Solution Design
Develops innovative and practical solutions to address algorithmic bias.
Exemplary
4 PointsProposes highly effective and innovative solutions to mitigate bias, demonstrating a deep understanding of algorithmic design and data manipulation.
Proficient
3 PointsProposes practical and feasible solutions to mitigate bias based on a sound understanding of the problem.
Developing
2 PointsProposes solutions that are partially effective or lack feasibility.
Beginning
1 PointsUnable to develop meaningful strategies to mitigate bias.
Implementation and Evaluation
Implements proposed solutions and evaluates their impact on fairness.
Exemplary
4 PointsDemonstrates exceptional skill in implementing bias mitigation strategies and rigorously evaluates their impact, providing clear evidence of improved fairness.
Proficient
3 PointsSuccessfully implements bias mitigation strategies and provides a reasonable evaluation of their effectiveness.
Developing
2 PointsAttempts to implement mitigation strategies but faces challenges or provides a basic evaluation.
Beginning
1 PointsFails to implement mitigation strategies or evaluate their impact.
Ethical Considerations
Demonstrates awareness of the ethical implications of AI development and deployment.Ethical Analysis
Critically examines the ethical considerations surrounding AI, including privacy, accountability, and transparency.
Exemplary
4 PointsProvides a profound and insightful analysis of the ethical challenges posed by AI, demonstrating a sophisticated understanding of the complexities involved.
Proficient
3 PointsAnalyzes the ethical considerations surrounding AI and provides well-reasoned arguments.
Developing
2 PointsIdentifies some ethical considerations but provides simple analysis or justification.
Beginning
1 PointsShows almost no awareness of the ethical implications of AI.
Responsible AI Advocacy
Advocates for the responsible and ethical use of AI in society.
Exemplary
4 PointsDemonstrates a strong commitment to advocating for responsible AI practices and effectively communicates the importance of ethical considerations to diverse audiences.
Proficient
3 PointsArticulates the importance of responsible AI and advocates for ethical considerations in its development and use.
Developing
2 PointsExpresses basic support for responsible AI but provides little justification or advocacy.
Beginning
1 PointsShows no engagement with the concept of responsible AI.
Communication
Effectively communicates findings, recommendations, and ethical considerations related to algorithmic bias.Clarity and Organization
Presents information in a clear, concise, and well-organized manner.
Exemplary
4 PointsCommunicates complex information with exceptional clarity and organization, using compelling visuals and engaging language.
Proficient
3 PointsPresents information clearly and logically, with a well-organized structure.
Developing
2 PointsPresents information in a disorganized or unclear manner.
Beginning
1 PointsUnable to communicate information effectively due to lack of clarity and organization.
Evidence and Justification
Supports claims with relevant evidence and logical reasoning.
Exemplary
4 PointsProvides compelling evidence and persuasive reasoning to support all claims, demonstrating a mastery of argumentation.
Proficient
3 PointsSupports claims with relevant evidence and provides logical reasoning.
Developing
2 PointsProvides limited evidence and weak justification or logical reasoning.
Beginning
1 PointsFails to provide evidence or justification for claims.