Embedding inclusivity and support in foundational software teaching

Dr Emma Rand is a Senior Lecturer in the Department of Biology at the University of York and is the Biosciences Computational Skills Coordinator responsible for data science components of 140 credits of taught postgraduate (PGT) and undergraduate (UG) modules for >1000 students and involving 25 members of staff. The department has five “Becoming a Bioscientist” modules which comprise one third of the Department of Biology’s teaching in stages 1, 2 and 3 and are core to all programmes. Each one of these trains students in experimental design, research techniques and data analysis. The R language is introduced in the first semester to stage 1 students, most of whom have no coding experience, limited experience with computing outside of tablets and mobile phones, and considerable apprehension of the topic. Emma has pioneered approaches that help put students at ease and encourages effective learning. Emma herself left school at 16 and came to University as a mature student via an access course and this experience has informed her approach. She aims to make data science teaching as open and accessible as possible. This case study explains some of the approaches she uses.

At the start of each course the students are introduced to a “code of conduct” which sets the tone for the learning. As well as the usual reminders about what behaviours are accepted, the code of conduct also outlines the approach to learning, including encouraging the students to ask questions and highlights the collective learning of the whole cohort. The code of conduct also aims to normalise mistakes - students are often worried to try coding solutions in case they are incorrect, but making mistakes can be a vital path to learning. Inspired by The Carpentries, sessions include “live coding” where an instructor types out codes in real time and students follow along. The reality of having to type out the code tends to slow the instructor down, it also allows the instructor to highlight and fix the inevitable errors that occur (even the occasional deliberate one!) and which reinforces the importance of making mistakes and learning through experimentation.

One key component of inclusive learning is offering multiple routes to catch up if a learner falls behind. In the sessions, the instructors try not to place too much focus on how far students have progressed through the material, but rather on the time, effort and understanding. Students that miss sessions are encouraged to just pick up where they left off, to avoid the feeling of exclusion when they don’t know the material. In addition, Emma makes sure to highlight that doing something is better than nothing; it is important to celebrate small steps as a way to encourage learners that are at risk of dropping out, to enable them to participate and be proud of their learning. Students can get help during the sessions from instructors (including student helpers) and between sessions they can submit questions anonymously to a public board where both the question and answer are visible to all students. The anonymity removes the fear of “asking a stupid question” which again can be used as a route to catch up if a student feels they are falling behind.

In addition all students are provided with a full worked solution to all the problems set, so they can use this to refer to. Rather than viewing the solutions as a “cheat sheet” Emma commented that this is similar to how many people code - most people writing code would search the internet when they get stuck, and would learn from the solutions, so full worked answers are an effective learning tool.

Emma has noticed that students are increasingly entering University with little awareness of the wider research practices; some students are familiar with using a tablet but not a laptop, or have used the cloud to store data but haven’t ever created a system of folders to store files in a systematic manner. To address this the course doesn’t just teach the mechanics of the R language, but instead teaches the entire research workflow, from setting up a directory structure, collecting data, writing code to analyse the data and, of course, documenting the code. During their course student design and carry out an experiment, they then need to write code to analyse the data and submit that along with corresponding documentation for assessment. The assessment criteria for the course focus less on whether the code is “correct”, but more on the workflow, reproducibility and documentation.

Data analysis and computation skills are assessed with the “Becoming a Bioscientist” modules an innovative and authentic way. Students produce a typical scientific report and are also required to submit “research compendia” including data and well-commented code that allows analyses and figures described in their report to be reproduced. This approach values transparency, openness and organisation over code “correctness” and allows students to more easily demonstrate what they have learned and gain credit for the problem solving process itself.

In the future, Emma would like all domains to include foundational research software skills in their undergraduate curricula but until that happens, all Biology students at York have the opportunity to learn vital software skills in an open and welcoming environment.