How can we design inclusive and accessible curricula for computer science?

After a brief hiatus over the Easter period, we are excited to be back with our series of online research seminars focused on diversity and inclusion, where in partnership with the Royal Academy of Engineering, we host researchers from the UK and USA. By diversity, we mean any dimension that can be used to differentiate groups and people from one another. This might be, for example, age, gender, socio-economic status, disability, ethnicity, religion, nationality, or sexuality. The aim of inclusion is to embrace all people irrespective of difference.

Maya Israel

This month we welcomed Dr Maya Israel, who heads the Creative Technology Research Lab at the University of Florida. She spoke to us about designing inclusive learning experiences in computer science (CS) that cater for learners with a wide range of educational needs.

Underrepresentation of computer science students with additional needs

Maya introduced her work by explaining that the primary goal of her research is to “increase access to CS education for students with disabilities and others at risk for academic failure”. To illustrate this, she shared some preliminary findings (paper in preparation) from the analysis of data from one US school district.

A computing classroom filled with learners.
By designing activities that support students with additional educational needs, we can improve the understanding and proficiency of all of our students.

Her results showed that only around 22–25% of elementary school students with additional needs (including students with learning disabilities, speech or language impairments, emotional disturbances, or learners on the autistic spectrum) accessed CS classes. Even more worryingly, by high school only 5–7% of students with additional needs accessed CS classes (for students on the autistic spectrum the decline in access was less steep, to around 12%).

Maya made the important point that many educators and school leaders may ascribe this lack of representation to students’ disabilities being a barrier to success, rather than to the design of curricula and instruction methods being a barrier to these students accessing and succeeding in CS education.

What barriers to inclusion are there for students with additional needs?

Maya detailed the systems approach she uses in her work to think about external barriers to inclusion in CS education:

  • At the classroom level — such as teachers’ understanding of learner variability and instructional approaches
  • At the school level — perhaps CS classes clash with additional classes that the learner requires for extra support with other subjects
  • At the systemic level — whether the tools and curricula in use are accessible

As an example, Maya pointed out that many of the programming platforms used in CS education are not fully accessible to all learners; each platform has unique accessibility issues.

A venn diagram illustrating that the work to increase access to CS education for students with disabilities and others at risk for academic failure cannot occur if we do not examine barriers to inclusion in a systematic way. The venn diagram consists of four fully overlapping circles. The outermost is represents systemic barriers. The next one represents school-level barriers. The third one represents classroom barriers. The innermost one represents the resulting limited inclusion.

This is not to say that students with additional needs have no internal barriers to succeeding in CS (these may include difficulties with understanding code, debugging, planning, and dealing with frustration). Maya told us about a study in which the researchers used the Collaborative Computing Observation Instrument (C-COI), which allows analysis of video footage recorded during collaborative programming exercises to identify student challenges and strategies. The study found various strategies for debugging and highlighted a particular need for supporting students in transitioning from a trial-and-error approach to more systematic testing. The C-COI has a lot of potential for understanding student-level barriers to learning, and it will also be able to give insight into the external barriers to inclusion.

Pathways to inclusion

Maya’s work has focused not only on identifying the problems with access, it also aims to develop solutions, which she terms pathways to inclusion. A standard approach to inclusion might involve designing curricula for the ‘average’ learner and then differentiating work for learners with additional needs. What is new and exciting about Maya’s approach is that it is based on the premise that there is no such person as an average learner, and rather that all learners have jagged profiles of strengths and weaknesses that contribute to their level of academic success.

In the seminar, Maya described ways in which CS curricula can be designed to be flexible and take into account the variability of all learners. To do this, she has been using the Universal Design for Learning (UDL) approach, adapting it specifically for CS and testing it in the classroom.

The three core concepts of Universal Design for Learning according to Maya Israel. 1, barriers exists in the learning environment. 2, variability is the norm, meaning learners have jagged learning profiles. 3, the goal is expert learning.

Why is Universal Design for Learning useful?

The UDL approach helps educators anticipate barriers to learning and plan activities to overcome them by focusing on providing different means of engagement, representation, and expression for learners in each lesson. Different types of activities are suggested to address each of these three areas. Maya and her team have adapted the general principles of UDL to a CS-specific context, providing teachers with clear checkpoints to consider when designing computing lessons; you can read more on this in this recent Hello World article.

Two young children code in Scratch on a laptop.

A practical UDL example Maya shared with us was using a series of scaffolded Scratch projects based on the ‘Use-Modify-Create’ approach. Students begin by playing and remixing code; then they try to debug the same program when it is not working; then they reconstruct code that has been deconstructed for the same program; and then finally, they try to expand the program to make the Scratch sprite do something of their choosing. All four Scratch project versions are available at the same time, so students can toggle between them as they learn. This helps them work more independently by reducing cognitive load and providing a range of scaffolded support.

This example illustrates that, by designing activities that support students with additional educational needs, we can improve the understanding and proficiency of all of our students.

Training teachers to support CS students with additional needs

Maya identified three groups of teachers who can benefit from training in either UDL or in supporting students with additional needs in CS:

  1. Special Education teachers who have knowledge of instructional strategies for students with additional needs but little experience/subject knowledge of computing
  2. Computing teachers who have subject knowledge but little experience of Special Education strategies
  3. Teachers who are new to computing and have little experience of Special Education

Maya and her team conducted research with all three of these teacher groups, where they provided professional development for the teachers with the aim to understand what elements of the training were most useful and important for teachers’ confidence and practice in supporting students with additional needs in CS. In this research project, they found that for the teachers, a key aspect of the training was having time to identify and discuss the barriers/challenges their students face, as well as potential strategies to overcome these. This process is a core element of the UDL approach, and may be very different to the standard method of planning lessons that teachers are used to.

A teacher attending Picademy teacher training laughs as she works through an activity.
Having time to identify and discuss the barriers/challenges students face, as well as potential strategies to overcome these, is key for teachers to design accessible curricula.

Another study by Maya’s team showed that an understanding of UDL in the context of CS was a key predictor of teacher confidence in teaching CS to students with additional needs (along with the number years spent teaching CS, and general confidence in teaching CS). Maya therefore believes that focusing on teachers’ understanding of the UDL approach and how they can apply it in CS will be the most important part of their future professional development training.

Final thoughts

Maya talked to us about the importance of intersectionality in supporting students who are learning CS, which aligns with a previous seminar given by Jakita O. Thomas. Specifically, Maya identified that UDL should fit into a wider approach of Intersectional Inclusive Computer Science Education, which encompasses UDL, culturally relevant and sustaining pedagogy, and translanguaging pedagogy/multilingual education. We hope to learn more about this topic area in upcoming seminars in our current series.

Four key takeaways from Maya Israel's research seminar: 1, include students with disabilities in K-12 CS education. They will succeed when given accessible, engaging activities. 2, consider goals, anticipated barriers, and the UDL principles when designing instructions for all learners. 3, disaggregate your data to see who is meeting instructional goals and who is not. 4, share successes of students with disabilities in CS education so we can start shifting the discourse to better inclusion.

You can download Maya’s presentation slides now, and watch the seminar recording here:

Attend the next online research seminar

The next seminar in the diversity and inclusion series will take place on Tuesday 4 May at 17:00–18:30 BST / 12:00–13:30 EDT / 9:00–10:30 PDT / 18:00–19:30 CEST. You’ll hear from Dr Cecily Morrison (Microsoft Research) about her research into computing for learners with visual impairments.

To join this free event, click below and sign up with your name and email address:

We’ll send you the link and instructions. See you there!

This was our 15th research seminar — you can find all the related blog posts here.



A worthy goal but most US school districts need to teach reading, writing and arithmetic as US students are almost illiterate in many cities. I’m astounded by how many young people I meet today that can barely read and write.


First thing they should do is remove the subjects that are not even related to computer science. Some schools in US mix up mech./electrical subjects in CS course and make it so confusing.


The UDL paradigm represents, inmy opinion, opportunity to alter the entire curriculum (not just computer science) to adapt the learning process to optimally utilize the strengths (and possibly ameliorate the weaknesses) in each learner’s capabilities profile (not just learning). I believe that courseware could be created and then disseminated (over the internet, but also on flash drives for locales without good internet access) that could be directed to individual learners whose needs would first be identified by testing and observation (also mediated over the internet or via flash drive). Note I say ‘learner’, not ‘child’–the challenges of the coming era will require many, if not most, persons to engage in continuous or at least periodic education. The sort of research described here, if properly supported, could potentially provide gains in educational productivity similar to those that the ‘green revolution’ did for agriculture. At a time when the US education system is, in my own estimation, verging on a state of utter collapse, pursuing new methodology appears utterly essential.,


Engaging activities is definitely a good way to generate better interest. Learning CS is now definitely simpler thanks to so many advancement in this field. Focusing on understanding concepts rather than just learning languages and algorithms may yield better results

Leave a Comment

Comments are closed