Designing inclusive class plans for cohorts with a high level of differentiation, with particular attention to accessibility needs of students who are neurodivergent or have learning differences

Please note: I’m submitting this report in place of the blog posts for the unit. It covers the same topics as requested in blog posts, including research methods, related work, and the research design and motivation.

Introduction

The project is closely related to my teaching practice as an Associate Lecturer at UAL Creative Computing Institute. I teach mostly subjects related to creative coding, which is programming with a creative application. As a transdisciplinary subject, the courses in Creative Computing attract students from a range of different backgrounds, including creative disciplines, technical disciplines, and people from different backgrounds looking for a change in their educational or career path. This creates a unique and inspiring environment to be and work in, but comes with a lot of challenges regarding differentiation in student cohorts, particularly in terms of technical and programming knowledge.

For the past 2.5 years at CCI, I have taught on a number of both undergraduate and postgraduate courses. This study focuses on the experiences of Foundation students. Last term, I had the chance to re-design a module being their first class in Creative Coding, called Foundational Methods for Creative Computing One. I also prepared all of the class materials. This project focuses on evaluating this effort and finding ways for improvement.

Class description and structure

The class focuses on equipping the students with a good understanding of basic programming concepts, such as variables, operators, conditional statements, loops, functions, etc. Although all of the learning outcomes refer to the Knowledge marking criterion (University of the Arts London, no date), typically referring to technical skills, in the labs and homework tasks I ask students to apply the newly learnt programming concepts to short creative tasks. For the final assessment, students need to submit links to 1 piece of work for each week. For example, here’s a lab task from “Week 05: Algorithmic Patterns | Binary systems and the Modulo Operator”:

In pairs: Have a look at the code below (examples from last week). Try to make a pattern using a central composition (based on a circle) and the modulo operator. You can use sketches from Week 4 and this week as a starter.

The class structure consists of a lecture with an integrated live coding demonstration lasing about 1.5 hrs in total, about 1.5 hours for labs, and 30 mins for drop-in tutorials.

Project rationale

As the student groups I work with have a lot of differentiation in terms of technical background, the project aimed to provide a better understanding of the student experience and ensure no-one feels left behind.

The project incorporates an accessibility aspect, and dedicates space in the study and planned interventions to students with accessibility needs. As someone who’s neurodivergent and was struggling with health issues during my 2 masters degrees, this is also of personal importance to me.

As an AL completing the PGCert, I was looking for a topic that would closely align with my teaching practice due to time limitations.

I also wanted the topic to help my professional development and future work, as well as benefit the students in longer term. As I will be teaching a continuation of the researched unit in Block 2, the same group of students will be able to benefit from the introduced improvements, which was an important aspect of ethical research design.

Background

Several authors prove the preference of live coding demos in early stages of programming education as opposed to working with static code. It has also been proven that the first approach produces learning results at least as good as the latter. It is not clear whether the live coding method may produce better outcomes (Raj et al., 2020; Selvaraj et al. 2021; Anindyaputri et al. 2020).

Also project-based learning and the “learning-by-making” approach is well established in the field (Papert, 1991). This constructionist approach allows the students for readily applying the new knowledge to a real-life context, as well as collaborate, learn from each other, and explore a creative topic of their own choosing.

In the constructivist concept of scaffolding, the tutors support students in manners such as splitting tasks into smaller pieces and designing the learning process in a way that each task builds on the previous knowledge (Wood, Bruner and Ross, 1976).

Methodology and methods

The project deployed the Participatory Action Research methodology (MacDonald, 2012), which drove the research design and choice of methods. Data collection adopted a mixed-methods approach and was split into 2 parts. The first part was a mixed-method anonymous questionnaire (Leavy, 2017). The full questionnaire can be found in the previous post. This method was selected with the intention of gathering honest student feedback, with anonymity at the centre for both ethical and practical reasons. As I will mark the unit’s final submission and continue to teach the students during the course, they could feel uncomfortable with providing honest feedback non-anonymously.

The questionnaire data was analysed using simple statistical methods for quantitative data and a relaxed approach to reflexive thematic analysis for qualitative data (Clarke and Braun, 2016), looking for patterns based on open coding in the qualitative data.

Initially, a second survey was planned at the end of the term, but there were not enough classes left with a “standard structure” to fully evaluate the introduced changes.

The second data collection method was non-intrusive observation, in line with the PAR methodology, acknowledging my positionality and bias as a teacher.

Study

The data collection via the questionnaire took place in class during a dedicated time. The survey was a part of gathering general feedback about the class. Students who were 18 or older could provide consent for their answers to be used for research. The Ethical Research Plan and the Participant Information Sheet are linked in the previous post.

Out of 13 students, 10 filled the questionnaire, and 9 students who were 18 or older gave permission to have their data used in the research project.

After a 2-week break for data analysis, I started introducing changes in the classroom. These were being evaluated via non-intrusive observation and documented using notes, ensuring to personal or identifying data about the students were collected.

Key findings: Quantitative data

The students were asked to rate the difficulty of the class material on a 5-point Likert scale. 7 out of 9 participants found the class material “The right amount of challenging”. The other 2 participants said the class material was “Somewhat too challenging”. There were no answers stating the class material “Way too challenging”, “Could be a little more challenging”, or “Should be much more challenging”.

I am happy with these answers as a sign of the right difficulty level of the class. Both participants who selected the class material is “Slightly too challenging”, also declared they spend 2-4 hours per week on self-directed learning. Based on credits, my expectation for weekly self-directed learning for this class would be around 6 hours.

Key findings: Qualitative data

Here are the key changes in the classes that the students requested:

1. Clearer instructions, tasks split into smaller bits.
2. Space for the students to learn at their own pace during the classes.
3. The coding demos should be at a slightly slower pace.
4. For the code to be provided before the live coding demo.
5. Some students asked for more tasks to work with ready code.
6. More diverse materials on Moodle including videos and reading.
7. More advanced homework options.

Other feedback:

1. The live coding demos received positive feedback.
2. Students notice the following classes and tasks within a class build on one another.
3. Some students struggle with language, which creates additional learning difficulties for them.

Accessibility-related feedback:

1. Only one student reported accessibility needs, which were the following:

I am neurodiverse, having ADHD and Autism, so having a slower pace within working in lesson would be better, still retain the same working relationship as before so as to not make me feel alienated from my peers

The same person requested for the code from live demos to be shared during the class rather than after, and for the coding demos to be slower.

Introduced interventions

The following interventions were introduced to further classes:

1. I started splitting technical tasks into smaller bits with clearer instructions, in line with the concept of scaffolding.
2. I started giving providing optional prompts for creative tasks, for example:
   
   Optional prompts [for a p5.js game]:
   • You’re a player running away from a giant space worm.
   • You’re a zombie and you need to collect brains to survive.
   • You’re a unicorn jumping from one rainbow to another.
   • You collect flowers of different kinds to complete a predefined bouquet.
     
3. I tried to introduce more project-based learning, and gave the students who needed to complete the tasks at their own pace an option to participate based on a ready template.
   Template code below. Sample outputs from this class – laser-cut winter decorations – can be found in the previous post and in the slides. https://editor.p5js.org/marysiatanska/sketches/-ZFXFYbUx
4. I slowed down during the coding demos and started checking in with students more frequently.
5. I started sharing links to my live coding demos and saving them frequently so that the students can go to the code and catch up if they need to; they don’t have to wait till the end of the class. The original request was to share the ready code in advance. I do not agree with this firstly as I’d expect less attention in class if they have access to ready code; secondly, because the code from live demos is often partially improvised or takes in prompts from students in the classroom; having a different pre-shared version may increase the confusion.
6. In response to the students’ feedback, I started introducing more diverse tasks as opposed to doing a coding demo from scratch. This allows students to complete more advanced projects. Examples:
   
   "understand, comment on and build on a piece of code"
"base your sketch on this template"

Discussion

Overall, the students gave me good advice on how to improve the class. A lot of the student feedback is in line with existing literature, particularly on the following topics:

• constructivist scaffolding (Wood, Bruner and Ross, 1976). In line with student feedback, I started splitting tasks into smaller fragments with gradually increasing difficulty level.
• constructionist “learning-by-making” (Papert, 1991). The students seem to enjoy project-based learning. The introduced changes allowing them to work with an existing codebase allows for more .
• literature on live coding demos in early stages of programming education (Raj et al., 2020; Selvaraj et al. 2021; Anindyaputri et al. 2020). Students enjoy live coding demos; it’s not clear whether it improves learning outcomes. This would need further work. Students prefer to work with varied methods and including ready code fragments into the classes as it allows for participation of students who learn at a slower pace and allows for more advanced projects. However, this approach might not be practical for the first few sessions, when the students need to understand and practise basic programming concepts.

Limitations and Further Work

After the survey, we didn’t have many classes with “standard structure” left in the same block, which provided limited opportunities for evaluating the interventions so far. I will be teaching a continuation of this class next block and intend to further apply the feedback, as well as conduct another iteration of the survey.

The study was conducted before the final submission and assessment for the unit. I believe I might be able to better assess the project results after the unit assessment.

One of the students mentioned they struggled with the language barrier, a problem well-known throughout UAL. I believe this is a topic for a whole new project.

A lot of the student feedback, including splitting tasks into smaller bits or adding more resources on Moodle, was my intuition before the study. It was, however, difficult to implement on an AL contract due to time constraints, particularly if I’m developing all contents from scratch.

References

Anindyaputri, N.A., Yuana, R.A., and Hatta, P. (2020). Enhancing Students’ Ability in Learning Process of Programming Language using Adaptive Learning Systems: A Literature Review. In Open Eng. 2020; 10, pp. 820–829.

Clarke, V., Braun, V. (2016). Thematic Analysis. In: Lyons, E., Coyle, A. (eds.) Analysing Qualitative Data in Psychology, 2nd edn., pp. 84–103. London, UK: Sage Publications.

Leavy, P. (2017). Research design: quantitative, qualitative, mixed methods, arts-based, and community-based participatory research approaches. New York, NY: Guilford Press.

MacDonald, C. (2012). Understanding Participatory Action Research: A Qualitative Research Methodology Option. In: Canadian Journal of Action Research Volume 13, Issue 2, 2012, pp. 34-50.

Papert, S. (1991). Situating Constructionism. In; Seymour, I. and Papert, S. (reds). Constructionism. Norwood, NJ: Ablex Publishing Corporation.

Raj, A.G.S., Gu, P., Zhang, E., R, A.X.A., Williams, J., Halvevrson, R., Patel, J.M. (2020). Live-coding vs Static Code Examples: Which is better with respect to Student Learning and Cognitive Load?, in: ACE’20: Proceedings of the Twenty-Second Australasian Computing Education Conference. Available at: https://doi.org/10.1145/3373165.3373182 (Accessed: 23.12.2024).

Selvaraj, A., Zhang, E. Porter, L., Raj., A.G.S. (2021). Live Coding: A Review of the Literature, In: ITiCSE 2021, June 26-July 1, 2021, Virtual Event, Germany. Available at: https://dl.acm.org/doi/10.1145/3430665.3456382 (accessed: 23.12.2024).

University of the Arts London (no date). Assessment and marking criteria. Available at: https://www.arts.ac.uk/study-at-ual/course-regulations/assessment (accessed: 29.01.2025).

Wood, D., Bruner, J.S., Ross, G. (1976). The Role of Tutoring in Problem Solving. In: J. Child. Psychol. Psychiar., vol. 17, pp. 89-100. Oxford, UK: Pergamon Press.