The ethics component addresses the issues of bias, data privacy, social and cultural implications, and fairness in AI. For each issue, we design discussion questions and class activities. Relevant reading and learning resources are also offered.

 

Social Role, Embedded Biased: A Reflective Statement

To put simply, self-reflexivity refers to how “who we are” influences “what we do”. In other words, our positionality including our beliefs, values, perceptions, and assumptions explicitly and/or implicitly shapes our actions in a context and consequently brings certain outcomes. The purpose of this assignment is to make us become more conscious of our individual, social, and cultural identities, and how our positionality influences our choices of use cases or projects.

Discussion questions:

The discussion questions lead to the activity, but the instructor can modify and adapt any parts of the suggested discussion questions, activities, or reading.

Activity

Please write a paragraph that briefly describes yourself and your identity, and how your identity influences your choices of courses, class projects, etc..

Suggested reading

How to Write a Reflexivity Statement (for professional or personal purposes)

Implicit bias test

Data Management Issues/Privacy: Research Ethics

 The goal of data management is to ensure that data collection and modeling is conducted ethically. Data mining often uses human behavioral data regulated by different legal regimes, like the HIPPA Privacy Rule and the FTC Act. In addition, research involving human subjects needs to be reviewed and approved by IRB (the Institutional Review Board). All of these legal and regulatory protections aim to protect privacy and confidentiality of participants. Respecting human participants who willingly provide data is another important dimension of research ethics.

Discussion questions

The discussion questions lead to the activity, but the instructor can modify and adapt any parts of the suggested discussion questions, activities, or reading.

Activity

Please write a data management plan that ensures privacy and confidentiality, shows respect to human participants, and minimizes potential risks and harm, in data collection, data storage, data analysis, and data reporting.

Suggested reading

Social and Cultural Implications: The So-What Question

 Data is a means to knowledge. The ultimate goal of data science is knowledge generation and decision making. Computer algorithms are powerful data engineering tools, but we, as human agents, need to answer the so-what questions. Data visualization is very helpful for interpretation of findings. The meaningfulness of data research and applications depends upon how well one can use his or her domain knowledge (like journalism, advertising, health communication, marketing, geography, economics, etc.) to interpret data and models.

Discussion questions

The discussion questions lead to the activity, but the instructor can modify and adapt any parts of the suggested discussion questions, activities, or reading.

Activity

Please write 2 to 3 paragraphs to answer the so-what question of your project/application: What is the broader social impact of the application? Did the data/model/application provide new insights and evidence to an issue at focus? How would your data/model/application bring awareness of the issue to the public? What do you think the policy makers should do to resolve the issue? Did the data/model/application and visualization reveal anything that you didn’t know or notice before about the social issue?

Suggested reading

Fairness in AI: Fairness Audit

AI techniques using big data and algorithmic processing are increasingly used to guide important social decisions, including hiring, admissions, loan granting, and crime prediction. However, AI is just as fair as the data, and the data are gathered from human activities. Data is often biased; data are as biased and flawed as human beings. While we assume that machines are neutral, there is evidence that algorithms may sometimes learn human biases and discrimination from data, rather than mitigating them.

Discussion questions

The discussion questions lead to the activity, but the instructor can modify and adapt any parts of the suggested discussion questions, activities, or reading.

Activity

The goal of this activity is to conduct a fairness audit to evaluate the data/model/application. For example, you can create audit metrics and conduct evaluation. Or, you can write a critical analysis of the data/model/application using critical studies theories, like feminism, social semiotics, etc. You can also create your own way of evaluating fairness.

Suggested reading

Welcome Back!

Login to your account below

Create New Account!

Fill the forms bellow to register

Retrieve your password

Please enter your username or email address to reset your password.