Advancing educational equity with UDL and generative AI

Key points:

As we all struggle along the path toward true educational inclusion, we face four pillars of equity, as described Rochelle Guiterize: Access, Success, Power and Identity.

Educators concerned about equity often excel in access. Opening the doors to all students is an obvious move. However, we must continually push systems forward so that all students can succeed in spaces where they have ownership and feel a sense of belonging (identity). Otherwise, equity and inclusion remain just a dream.

While we recognize that some of these elements require major systems changes, we also want to challenge all computer science teachers to lead by example. Moving computing, with its long history of exclusion, into an inclusive future will cause ripple effects across all content areas. . Using the AiiCE Principlesthat recommend adopting approaches that are responsive to students' identities (Alliance for Identity-Inclusive Computing Education, 2023), we will suggest steps toward an inclusive education pedagogy with Universal Design for Learning (UDL) and generative ai thinking partners.

A first step towards inclusive education can be taken by adopting UDL. According to the CSTA: Inclusive Teaching Pedagogies, “UDL is an instructional planning approach designed to provide all students with an equal opportunity to learn by removing barriers that prevent students from fully participating in their classroom communities” (White, 2023). However, this is a time-consuming (although worthwhile) task for teachers who are already taxed.

Within the framework of working smarter, not harder, we will outline a way to begin integrating UDL principles into lessons, moving toward greater equity and inclusion through the use of Generative ai (GenAI) tools. The generative model used is ChatGPT 3.5 (ai?utm_source=substack&publication_id=1180644&post_id=139374255&utm_medium=email&utm_content=share&utm_campaign=email-share&triggerShare=true&isFreemail=true&r=2urutn” target=”_blank” rel=”noreferrer noopener”>For optimal use we recommend ChatGPT 4).

Teaching the average student has never been effective. Our students have a wide range of different brains, with different sensory and processing capabilities. Good teachers are finding ways to meet the learning needs of all these diverse brains within the same class.

UDL uses the foundations of neuroscience to provide educators with a frame empower all students (CAST, 2018). UDL is a process, not a product, and requires teachers to reconsider their planning and delivery of instruction. While this doesn't necessarily mean asking teachers to do more, it absolutely does ask them to do something different. As teachers strive to transform their teaching practice, generative ai offers strong opportunities. When we combine a proven, research-based framework like UDL with ai, we get one step closer to the goal of true inclusion of all students in computer science classes.

Implementing UDL requires rethinking lesson development and planning. Ralabate (2016) asks us five fundamental questions that allow teachers to begin to transform their practice. As teachers embrace this transformation, generative ai can be a thought partner to use the five fundamental questions efficiently. These questions revolve around the accessibility, flexibility, freedom from bias, validity and reliability of our learning activities.

Below we address the first four of these questions, along with generative ai prompts that can be used to increase the speed of implementation of each of these questions.

The last question revolves around reliability. Reliability measures the ability of a learning activity to achieve its objectives. Is the variability in my student's performance entirely due to her performance, or is there variation that is due to the design of the activity (Ralabate, 2016)? It is impossible to truly eliminate variation due to design, but it will be minimized if the first four questions are carefully considered and implemented in the design process. As a final reliability check, GenAI can be used to triangulate scores: ask it to evaluate student data against a rubric. By comparing multiple GenAI responses to teacher results, we can minimize implicit bias and ensure that the grades we give are authentic measures of student learning.

Systems produce what they were designed to do. Our educational system was built to produce inequitable results, and that is what it produces. We believe that computer science educators can rise to the current challenge and remake their teaching in a way that effectively educates all brains—brains that have extremely diverse needs. We know the why (equity), we know the how (UDL) and with generative ai, we now have the means to achieve what is demanded at the moment.

References

Alliance for Identity-Inclusive Computing Education (2023). AIICE IIC Principles. https://identityincs.org/resources/aiice-iic-tenets/

CAST (2018). UDL and the learning brain. Wakefield, MA. Retrieved from http://www.cast.org/products-services/resources/2018/udl-learning-brain-neuroscience

Gutiérrez, R. (2011). Context matters: How should we conceptualize equity in mathematics education? In Equity in discourse for mathematics education: Theories, practices, and policies (pp. 17-33). Dordrecht: Springer Netherlands.

Ralabate P. (2016). Your UDL Lesson Planner: The step-by-step guide to teaching every student. Brookes Publishing.

White, S.V. et al. (2023, June 5). Inclusive teaching pedagogies. Association of Computer Science Teachers. https://csteachers.org/inclusive-teaching-pedagogies/

Biographies: