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 AI 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 school 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) demonstrating an 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 year 3 implementation of this new curriculum focused on social justice in computing. Results of 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 the majority of prior research had focused on using existing CS curricula and programming environments to meet the needs of students with disabilities, this study shows that targeted curricula that are 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, and use CS as a tool for civic engagement. This is the exact kind of thing Yadav and Heath argued for in their paper, which focused on justice-oriented computer science education. Specifically, the authors argued that CS needs to center criticality that prepares students "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 were 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 courses.

The Changing Expectations CSforALL RPP has been successful in cultivating engagement, identity, and a sense of belonging in computing for African American and Hispanic students with disabilities. 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 and creating diverse representation among AI developers, support must be made available from AI tech corporations and K12 education resources.

For actionable steps towards change, please see A Guide for Educators, Families, and Communities to Advance Positive Outcomes for Black Students with Disabilities in STEM and Computer Science.

See the Changing Expectations CSforALL RPP’s research document "Identifying Systemic Barriers: Computer Science District and School Policies for African American and Hispanic Students with Disabilities" as a reference for more information and examples.

Use the Changing Expectations GPT for Inclusive Pathways in Computer Science and Artificial Intelligence and Broadening Participation in CS and AI Education.

The Broadening Participation in Computer Science and AI Education GPT is designed to support educators, policymakers, and stakeholders in addressing the systemic inequities faced by underrepresented students, particularly Black and Hispanic students with disabilities, in computer science (CS) and artificial intelligence (AI) education. By leveraging evidence-based practices and cutting-edge technology, the tool empowers schools and districts to:

Identify Systemic Barriers: Analyze district and school-level policies that limit access to CS education for Black and Hispanic students with disabilities, as outlined by the African American and Hispanic Students with Disabilities in Computer Science Research Alliance study.

Promote Inclusion and Accessibility: Implement Universal Design for Learning (UDL) strategies and culturally relevant pedagogy to make computer science education accessible to all students.

Integrate AI-Powered Solutions: Use AI-driven voice-enabled chatbots to engage students with disabilities in CS learning, encouraging participation through culturally relevant, personalized, interactive, and accessible experiences.

Encourage Role Models and Representation: Highlight the importance of diverse mentors and peers representing the students’ cultures in CS classrooms to inspire students and create pathways to success.

Enhance Educator Training: Provide resources to prepare teachers in inclusive CS and AI instruction, building their capacity to address the unique needs of Black and Hispanic students with disabilities while fostering a sense of belonging.

Whether you're an educator aiming to diversify computing and AI education, a policymaker seeking data-driven insights, or an advocate for equity in computing education, this custom GPT tool equips you with actionable strategies and resources to broaden participation of Black and Hispanic students with disabilities in the CS and AI education fields.

NSF Award 1923199 – Changing Expectations CSforALL Research-Practice Partnership (RPP)

Over six years, the Changing Expectations CSforALL Research-Practice Partnership (RPP) project expanded access to computer science (CS) and artificial intelligence (AI) education for Black and Hispanic students with disabilities. These students often face barriers to participation in advanced STEM education, including limited opportunities to explore emerging technologies like AI. This project addressed that gap by designing engaging, real-world learning experiences rooted in students’ interests and communities, while building capacity among the educators and institutions that support them.

From 2019 to 2025, the project team—made up of educators, researchers, and technologists—developed and implemented an AI voice chatbot design curriculum that encouraged students to use computing to solve problems that mattered to them. Students created AI voice-enabled chatbot tools to address topics such as mental health, school safety, and access to community services. This hands-on, design-based approach helped students develop technical and critical thinking skills, while also fostering their identity as creators of responsible AI technology. In parallel, the project provided professional development for teachers to lead AI development lessons and support inclusive implementation.

Intellectual Merit

• Advancing Knowledge in AI and CS Education: The project introduced a novel curriculum integrating voice-based AI chatbot development, emphasizing design-based learning and ethics. This contributed to new understanding of how emerging AI technologies can be introduced in inclusive, K–12 settings.

• Innovative Pedagogy and Research-Practice Collaboration: The project combined teacher practice, student feedback, and education research to co-develop tools and strategies for real-world AI instruction. This model helped bridge research and classroom implementation in meaningful ways.

• Focus on Ethics and Critical Thinking: Students engaged with algorithmic bias, voice interface design, and problem-solving in social contexts. These components supported broader research goals around ethics in computing and student-centered CS and AI learning.

• Contributions to Education Research: Field observations and teacher interviews yielded insights into barriers and opportunities in CS education for underrepresented learners with disabilities, informing broader research on inclusive CS pedagogy and accessibility.

Broader Impacts

• Broadening Participation in CS and AI: The project reached Black and Hispanic students with disabilities—groups historically underrepresented in computing—and demonstrated that inclusive, interest-driven learning increases participation and persistence.

• Improving Teacher Capacity and Tools: Through professional development and the creation of an AI-powered educator assistant, the project enhanced teacher readiness to deliver inclusive CS and AI instruction beyond the life of the grant.

• Supporting Community-Based Learning: Weekend and afterschool AI coding makerspaces provided students with hands-on experiences in nontraditional settings. These efforts engaged families, community mentors, and institutions of higher education to create stronger support networks for CS and AI learning.

• Developing Future Workforce Skills: Students built real-world skills in coding, responsible AI development, and design thinking. Their experiences increased confidence and career awareness, helping to prepare the future AI and tech workforce.

• Encouraging Systems-Level Change: The project’s findings contributed to conversations about how educational systems can better support access to computing. Districts and school leaders have expressed interest in adopting or expanding the model.

Conclusion

The Changing Expectations CSforALL RPP project demonstrates how artificial intelligence and computer science can be taught in ways that are inclusive, community-connected, and meaningful to students’ lives. It empowered young people to use technology to solve real-world problems, supported educators in building more accessible CS learning environments, and created a foundation for continued growth in CS and AI education. Its outcomes contribute not only to teaching and learning but to a broader vision of technology shaped by our voices and experiences.