Computer Science for ALL Research-Practitioner Partnership (CSforAll RPP)

Broadening Participation to Include African-American and Hispanic Students with Disabilities in Computer Science Learning Using Voice User Interface Project-Based Learning

The National Science Foundation awarded a grant to Changing Expectations for a Computer Science for ALL Research-Practice Partnership (CSforALL RPP).  In year one, the RPP was developed with education practitioners from six Central Texas secondary schools, policymakers, and researchers working together to solve problems of practice to create equity in computer science education. The project was designed to study and increase African American and Hispanic students with disabilities interest, engagement, learning, knowledge, sense of belonging, and intentions to persist in computer science education. The project conducted research through the African American and Hispanic Students with Disabilities in Computer Science Research Alliance, a working group consisting of computer science educators, STEM, CTE, Tech, and Special Educators at school sites in Texas school districts, researchers and evaluators. The project also implemented an artificial intelligence education curriculum for the teachers and students.

In year two, teachers at five middle and high school sites provided Saturday remote learning sessions to support the targeted students in learning how to create artificial intelligence (AI) voice-enabled chatbots using a project-based approach to solve culturally-relevant problems in student homes, schools, and communities. The project has also provided teacher professional development on the IBM Watson Assistant (i.e., AI chatbot), artificial intelligence, special education strategies, universal design, and project-based learning. As a result of this project, African American and Hispanic students with disabilities in secondary schools learned to be the creators of trustworthy artificial intelligence innovations for social justice, rather than just consumers of that technology.

In year three, we worked to identify solutions to broadening the participation of African American and Hispanic students with disabilities in computer science education. We also addressed the importance of elevating issues of racial equity and the intersectionality of race and disability for African American and Hispanic Students with disabilities. Computing and special education teachers at a couple of Texas schools sites and one high school in Miami, Florida again teamed up to implement after school and/or Saturday sessions to support the targeted students in learning how to create artificial intelligence (AI) voice chatbot projects for social justice. The project’s teachers and students earned IBM digital badges (micro credentials) that demonstrate understanding of creating chatbots by leveraging IBM Watson. Then, teams of students designed AI voice-enabled chatbot projects to solve student-selected social justice problems for their friends, families, and community members. The Design Justice principles provided guidance to create AI voice chatbot projects led by marginalized communities to dismantle structural inequality and to advance collective liberation and survival for African American and Hispanic people. During the final project year, we wrote a case study that examined how the curriculum that focused on social justice in artificial intelligence education can be used to educate Black and Hispanic students with disabilities.

Numerous African American and Hispanic Students with disabilities are confronted with systemic and policy-based challenges preventing access to K-12 STEM-related and computer science education. In addition, the African American and Hispanic Students with Disabilities in the Computer Science Research Alliance conducted an NSF-funded study to understand teachers’ perceptions of district and school policies and practices that hinder the participation of African American and Hispanic students with disabilities in computer science education in Central Texas. The project’s first research study fills a critical gap in the literature concerning the systemic barriers affecting African American and Hispanic students with disabilities in K12 computer science education.

Moreover, the Changing Expectations CSforALL RPP conducted a second research study to understand ways to broaden participation in computing by creating opportunities for African-American and Hispanic students with disabilities to learn to design AI voice chatbots for social justice. The 3-year implementation of this new curriculum focused on social justice in computing. Results for that NSF-funded research study suggest that both teachers and their students saw value in learning computer science through the AI chatbot for social justice course that was specifically designed for African American and Hispanic students with disabilities. Given that majority of prior research had focused on using existing CS curricula and programming environment to meet the needs of students with disabilities, this study shows that targeted curricula that is focused on social justice issues for students of color with disabilities could shift how teachers and their students engage in CS. One of the interesting findings of the study was that Mr. Wood believed that when it comes to social justice CS should highlight both algorithmic biases such as those present in facial recognition technologies as well as use CS as a tool for civic engagement. This is the exact kind of thing Yadav and Heath argued for in their paper focused on justice-oriented computer science education. Specifically, the authors argued that CS needs to center criticality that prepares students to "by interrogating the role of CS in the design and deployment of technologies that harm and oppress Black and Brown communities". Similarly, Yadav, Heath, and Hu suggested that we should prepare students to use CS as a tool for participation and change in their communities using citizen science practices. Another finding that emerged from this study was how Ms. Boone believed that CS provided students with disabilities with opportunities that are not afforded to them in any other class and increased their self-confidence. This is similar to the finding from Israel and colleagues who found that students with disabilities do not feel judged (by a computer) in CS classes in ways that humans may judge them in other classes.

The Changing Expectations CSforALL RPP has been successful in cultivating African American and Hispanic students with disabilities’ engagement, identity, and sense of belonging in computing. We also supported these students to develop competencies in inclusive and diverse artificial intelligence education. African-American and Hispanic students with disabilities attribute the curricula on designing AI voice chatbots for social justice and the learning environment to helping them grow as artificial intelligence learners and developers. However, for these interventions to be sustainable in diversifying the AI field to create diverse representation among AI developers, supports must not only be made available from AI tech corporations, but also from K12 education resources.

The Changing Expectations CSforALL RPP project has several outcomes and findings that address the intellectual merit and broader impacts as defined in the NSF merit review criteria.

Intellectual Merit:

1.     Advances Knowledge: The project contributes to advancing knowledge in computer science education by focusing on racial equity, disability, and social justice in CS learning. It shifts the focus from traditional computer science (CS) curricula to culturally responsive content that engages African American and Hispanic students with disabilities in designing AI voice chatbots for social justice. The project introduces an innovative approach to teaching computer science that emphasizes teaching and learning about critical analysis, algorithmic biases, and civic engagement, aligning with the principles of justice-oriented CS education.

2.     Novel Approaches: The project employs a unique approach to teaching AI and computer science to African American and Hispanic students with disabilities. Instead of relying solely on existing CS curricula, it creates and tailors AI education content to address the specific needs of underrepresented students, emphasizing their participation, empowerment, sense of belonging, and self-confidence. The integration of Design Justice principles adds a novel dimension to AI education, fostering a sense of agency and collective liberation among marginalized communities.

3.     Effective Pedagogy: The project's focus on project-based learning, teacher professional development, and collaboration with both educators and researchers enhances pedagogical practices. By involving teachers from various disciplines and locations, the project identifies effective strategies for engaging African American and Hispanic students with disabilities in computer science education, leading to improved learning outcomes and increased interest.

Broader Impacts:

1.     Diversity and Inclusion: The project addresses the underrepresentation of African American and Hispanic students with disabilities in computer science education. By creating targeted curricula and providing support for teachers, the project helps broaden participation and cultivate a sense of belonging among these students. This contributes to diversifying the AI field and ensuring that traditionally marginalized groups are included in the development of AI technologies.

2.     Educational Enhancement: The project enhances K-12 computer science education by offering specialized content that caters to the unique needs and aspirations of African American and Hispanic students with disabilities. Through the development of AI voice chatbot projects for social justice, students gain practical skills, critical thinking abilities, a sense of belonging, and confidence, positioning them for success in both CS and broader educational contexts.

3.     Workforce Development: The project contributes to the preparation of a diverse and inclusive future workforce in the AI and technology sectors. By empowering African American and Hispanic students with disabilities to become creators of AI technologies, rather than just consumers, the project equips them with valuable skills that can lead to future career opportunities and a more equitable AI landscape.

4.     Policy and Systemic Change: The project's research on systemic barriers affecting underrepresented students informs policy and practice. By conducting a research study on teachers' perceptions and identifying challenges hindering participation, the project provides insights that can drive systemic change within educational institutions, promoting equitable access to CS education for African American and Hispanic students with disabilities.

In summary, the Changing Expectations CSforALL RPP project's outcomes and findings align with the intellectual merit and broader impacts criteria outlined by the NSF. It advances knowledge, introduces novel approaches, and enhances pedagogical practices in computer science and artificial intelligence education. Simultaneously, the project promotes diversity, inclusion, educational enhancement, workforce development, and policy change, thereby contributing to a more equitable and inclusive AI field and society.