AI Literacy for Underserved Students: Leveraging Cultural Capital from Underserved Communities for AI Education Research

Research Background and Issues

• Issues and Challenges:
  1. Artificial Intelligence (AI) is profoundly impacting various sectors of society, and AI literacy education is becoming increasingly important for K-12 students.
  2. Current research and projects primarily focus on well-resourced schools and student groups, leading to the marginalization of underserved communities (constrained by socioeconomic status, gender, and race) in the field of AI literacy education.
  3. Underserved student groups lack access to AI education resources and opportunities, which are often influenced by structural inequalities, leaving them ill-equipped to critically reflect on the social and ethical issues of AI technologies.
• Significance:
  1. AI may inadvertently exacerbate historical biases in society, particularly stereotypes related to gender and race. Therefore, helping underserved students understand AI's potential impact on their communities carries significant social justice implications.
  2. These students possess unique cultural capital (such as distinctive skills and knowledge rooted in their community's history and culture), which, if effectively utilized, can enhance their engagement and outcomes in AI literacy learning.
• Research Motivation and Related Work:
  1. Critical pedagogy advocates for education that avoids rote knowledge transmission, instead integrating students' community backgrounds and experiences to cultivate critical social thinking.
  2. Previous AI education research has largely focused on simplifying complex technical issues, such as teaching tools and programming activities. However, educational approaches designed for underserved communities, especially those centered on cultural capital, remain relatively scarce.

Solution

• Methods and Innovations:
  1. Educational Framework Based on Cultural Capital Theory: Drawing on Tara Yosso's "Community Cultural Wealth" approach, identify the unique cultural capital of students in underserved communities and apply it to AI literacy education.
     • Identify three types of cultural capital: Resistant Capital, Communal Capital, and Creative Capital.
  2. Collaborative Learning Relationships: Emphasize collaboration between researchers and students to uncover students' unique abilities through dynamic dialogue, rather than having researchers dominate the educational process.
  3. Practical Activity Design: Develop hands-on activities involving AI classification and generation, coupled with discussions on AI ethics and social issues (e.g., anti-bias and misinformation) to enhance students' critical thinking skills.
• Implementation Steps:
  1. Background Investigation:
     • Conduct interviews to understand the community background, potential issues, and students' perceptions of AI.
  2. Development of Educational Materials:
     • Design modular courses including:
       • Conceptual teaching of core AI functions (classification, generation);
       • Practical applications relevant to students' communities, such as using AI to address issues like school bullying and healthcare access.
  3. Testing and Iteration:
     • Pilot the courses in public libraries across three underserved communities in the U.S., gradually refining the teaching materials and activity design.
  4. Analysis and Redesign:
     • Focus on analyzing how students' cultural capital facilitates their understanding of AI knowledge and connects to community issues.

Research Outcomes

• Specific Outcomes:
  1. Identification and Application of Three Types of Cultural Capital:
     • Resistant Capital: Students use their sensitivity to racial or gender inequalities to assess AI's racial and gender biases.
     • Communal Capital: Demonstrate concern for community needs by designing AI systems that benefit their social groups.
     • Creative Capital: Exhibit spontaneous innovation through unexpected operations (e.g., using AI classifiers to recognize faces) or analogies (e.g., comparing GANs to teacher-student relationships).
  2. Propose a new AI literacy education approach centered on cultural capital, leveraging students' lived experiences to stimulate critical reflection and enhance learning relevance.
• Advantages:
  • Compared to traditional one-size-fits-all AI education frameworks, this approach better integrates students' unique backgrounds and strengths, particularly focusing on underserved groups.
  • Moves beyond knowledge transmission to help students develop deeper social evaluation skills regarding AI systems.
• Experimental and Evaluation Results:
  1. All 26 participants (including economically disadvantaged, sexual minority, and racial minority students) completed four pilot teaching modules.
  2. Collaborative hands-on activities in the course design (e.g., DALL-E generation activities and community AI design projects) enhanced students' intuitive understanding of AI and its societal impact.
  3. Student-initiated new activities (e.g., using GANs to construct metaphors for teacher-student relationships) were incorporated into the improved teaching modules.
• Limitations and Future Directions:
  1. Limitations:
     • Comprehensive identification of students' cultural capital remains challenging, especially in the early stages of course design.
     • The small test scale requires larger-scale validation to expand the applicability of the approach.
  2. Future Directions:
     • Further refine the definition of cultural capital and its adaptability across different regions/countries.
     • Develop more sustainable and scalable course formats to standardize the mobilization of cultural capital.

Conclusion

This study provides a novel practical framework for AI literacy education for underserved students by identifying and mobilizing their Resistant Capital, Communal Capital, and Creative Capital. It enriches the integration of AI education with cultural capital theory and proposes significant improvements to educational methods from a social justice perspective. In the future, as the scale and diversity of implementation expand, this research approach has the potential to further promote educational equity in underserved communities.