Computing · JC 2 · Database Systems and Data Modeling · Semester 1

Ethical Use of Data

Students will discuss the ethical implications of collecting and using data, considering fairness, bias, and transparency.

• MOE Syllabus Outcomes
• MOE: Social Computing - Middle School

About This Topic

Ethical use of data requires attention to fairness, bias, and transparency throughout collection, storage, and application. JC 2 students in the Database Systems and Data Modeling unit discuss how poor practices lead to harm, such as discriminatory outcomes in algorithms or privacy violations. They address key questions: what ethical data use means, how bias arises, and scenarios raising concerns like surveillance without consent.

This topic supports MOE Social Computing standards by linking technical database skills to societal impacts. Students analyze real cases, including biased credit scoring or facial recognition errors affecting minorities, to develop judgment for responsible decision-making in computing careers.

Active learning excels with this abstract content. Role-plays of ethical dilemmas, group debates on bias mitigation, and collaborative case analyses help students confront trade-offs firsthand. These methods build empathy, sharpen arguments, and make principles stick through peer interaction and reflection.

Key Questions

1. What does it mean to use data ethically?
2. How can data be used unfairly or to create bias?
3. Discuss a situation where data collection might raise ethical concerns.

Learning Objectives

• Analyze case studies to identify instances of algorithmic bias in data collection and usage.
• Evaluate the ethical trade-offs between data privacy and the benefits of data-driven insights.
• Critique data models for potential sources of unfairness and lack of transparency.
• Propose mitigation strategies for ethical concerns related to data collection in specific scenarios.

Before You Start

Introduction to Databases

Why: Students need a foundational understanding of how data is stored and organized before discussing its ethical implications.

Basic Data Structures

Why: Understanding how data is structured helps students identify potential points where bias can be introduced or where transparency is lacking.

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.
Data Privacy	The protection of personal information from unauthorized access, use, disclosure, alteration, or destruction.
Transparency	The principle that data collection and usage practices should be clear, understandable, and accessible to those affected by them.
Fairness	Ensuring that data collection and algorithmic decision-making do not disproportionately disadvantage or discriminate against certain groups.

Watch Out for These Misconceptions

Common Misconception

Data is always objective and free of bias.

What to Teach Instead

Data inherits biases from collection methods, sources, and human choices. Group audits of datasets help students spot patterns like underrepresentation, fostering skills to question assumptions early.

Common Misconception

Ethical issues only arise from illegal data use.

What to Teach Instead

Fairness problems occur in legal practices, like targeted ads reinforcing stereotypes. Role-plays of everyday scenarios reveal subtle harms, encouraging nuanced ethical discussions.

Common Misconception

Transparency requires full public data access.

What to Teach Instead

Transparency means clear methods and limits, not total disclosure. Debates on risk-benefit balances clarify responsible sharing, building consensus through active peer exchange.

Active Learning Ideas

Debate Duos: Privacy vs Progress

Pairs research one side of a dilemma, such as mandatory health data sharing during pandemics. They debate against another pair for 10 minutes, then switch sides and reflect in writing. Class votes on strongest arguments.

45 min·Pairs

Case Study Stations: Bias Examples

Set up stations with cases like loan algorithm bias or social media targeting. Small groups spend 8 minutes per station noting ethical issues, proposed fixes, and database links. Groups share one insight per case.

50 min·Small Groups

Role-Play Rounds: Data Dilemmas

Assign roles like data analyst, citizen, and regulator in scenarios such as selling user data. Groups perform 5-minute skits, followed by whole-class debrief on fairness checks. Rotate roles for second round.

40 min·Small Groups

Bias Audit Challenge: Dataset Review

Individuals or pairs audit sample datasets for biases in gender or ethnicity representation. They document findings and suggest debiasing steps, then present to the class for feedback.

30 min·Pairs

Real-World Connections

• Tech companies like Google and Meta face scrutiny over how user data is collected and used for targeted advertising, raising questions about privacy and algorithmic bias in content recommendation.
• Financial institutions use algorithms for loan applications and credit scoring. Students can investigate how historical biases in lending data might perpetuate unfair outcomes for certain demographics.
• Law enforcement agencies exploring facial recognition technology must consider the ethical implications of its accuracy across different racial and gender groups, and the potential for misuse.

Assessment Ideas

Discussion Prompt

Present students with a scenario: A city wants to use public surveillance cameras with facial recognition to improve public safety. Ask: What are the potential ethical concerns regarding data privacy and fairness? What specific data points might be collected, and how could they be misused? What safeguards should be in place?

Quick Check

Provide students with short descriptions of two data collection projects (e.g., a health app tracking user activity, a social media platform analyzing user posts). Ask them to identify one potential ethical issue for each project and explain why it is a concern, referencing concepts like bias or transparency.

Exit Ticket

Ask students to write down one way data can be used unfairly and one concrete step a data scientist could take to promote fairness in their work. They should also define 'transparency' in their own words.

Frequently Asked Questions

What are main principles of ethical data use in computing?

Core principles include fairness to avoid discrimination, transparency in methods and decisions, and accountability for impacts. Students apply these by evaluating database designs for bias risks and privacy safeguards. In Singapore's context, align with PDPA guidelines to ensure compliance and trust in data systems.

How does bias creep into datasets and databases?

Bias enters via skewed sampling, like training AI on non-diverse data, or flawed modeling assumptions. In databases, poor normalization amplifies inequalities. Teach detection through hands-on audits where students query samples and visualize distributions to identify and mitigate issues.

How does active learning strengthen data ethics lessons?

Active methods like debates and role-plays immerse students in dilemmas, promoting empathy and critical thinking over rote memorization. Collaborative case studies reveal multiple viewpoints, while reflections solidify principles. These approaches make ethics relevant to real computing roles, boosting retention and application skills.

What real ethical concerns arise in data collection?

Concerns include consent gaps, such as apps tracking without clear notice, or aggregation creating profiles invading privacy. In Singapore, contact-tracing apps raised debates on data retention. Discuss via scenarios to weigh public good against individual rights, linking to database security features.