Focus on Social Impact
The data science and AI revolution is on a trajectory to further enhance inequity and injustice in society (Benjamin, 2020). It is our responsibility to train the next leaders of the data science workforce to ethically advance justice and accountability and understand the power that data science and AI can have on society. Integrating social impact into data science education affords the opportunity to offer authentic and relevant real-world problems that can foster in our students a sense of civic responsibility and inspire them to be agents of change. Social impact challenges are complex, often with no easy solutions. Addressing these challenges in the classroom can foster problem-solving skills, critical thinking, and the ability to evaluate issues from multiple perspectives. Ultimately, the goal is that students become motivated to engage in community projects and advocacy, and choose career paths where they can make a difference.
Opportunities for Teaching and Learning Social Impact Data Science
The work of social impact can be accomplished in many ways and you will see examples of this throughout the Playbook, including:
- Recruiting and supporting students from groups underrepresented in data science.
- Engaging communities in ways that empower the community to use their own data for purposes they co-define.
- Choosing course topics with titles and topics that directly point to social impact.
- Encouraging students to seek, explore, and analyze data sets that surface issues of social impact.
- At all stages of work, considering the potential social impact of that work.
- Building open science principles into courses.
- Incorporating accessibility practices into teaching and learning.
- Creating social mobility through educational pathways from community college through the Master’s or PhD programs.
- Designing educational experiences intentionally and collaboratively to build equitable and accessible opportunities and pathways in data science.
- Providing students with work-based learning and awareness of career opportunities.
Ethical Approach to Data Science
An important opportunity for teaching and learning social impact data science is normalizing ethical approaches to data science. Data sharing brings up questions of transparency, interpretation, and trust and is an important topic to address head-on with external partners and organizations. Projects should begin and end with an understanding of the needs and mission of project partners. Non-extractive practice and mutual benefit are core tenets of ethics in data science collaborations. Throughout the entirety of the data project lifecycle, it is important to keep the relevant legal, political, and cultural concerns in mind. Academics leading this work should work with partners to identify privacy and security concerns and ensure reasonable steps and mitigation measures are in place. It is also important to evaluate the appropriateness, feasibility, and sustainability of any technology implemented in projects with a default for open-source or low-code options whenever possible. Integrating ethically skillful practices in design and implementation is an ongoing process and should be reinforced throughout the project. Recognizing and managing not only the bias in the data but also possible unintended consequences, misuse, or misinterpretation is necessary.
Benjamin, R. (2020). Race after technology: Abolitionist tools for the new Jim code. Polity Books.