Joining the department of Computing a few years back, I was assigned to teach an undergraduate module called Audio-Visual Computing. Faced with problems in its original iteration, I reflected on the way I delivered the course and made adjustments which had a positive effect on the learning experience of the students. This essay narrates this story.
The course itself is a hybrid introduction to programming graphics and sound and it is a core part of the year one, term two computing curriculum. The cohort is quite diverse as it is made up by students from many undergraduate streams. There are students from the following BSc courses:
● Music Computing
● Digital Arts Computing
● Creative Computing
● Games Computing
● Business Computing
This diversity creates extra challenges as students come in with different appetites to embrace the material, different expectations and different technical abilities. Music Computing students show greater skill and interest at the audio part of the course, while Games Computing students for the graphics part. Both of these groups have usually some coding experience as hobbyists or by taking computing A-levels in school, while the Digital Arts Computing students usually come with A-levels in art and are usually not very numerate. Finally, Business Computing students don’t consider themselves artistic in any sense. They just want to become better general coders. The module has about 70 students, which is quite large for computing class. It was made up of a weekly 90 minute lecture followed by a 90 minute lab. In the lab part I was helped by two lab assistants.
The syllabus by its nature is challenging, but the mode of its delivery, its content “served” to a young, diverse cohort led to problems in my first year of running it.
Initial reflection and diagnosis
Having picked up on this module in my first year on very short notice, I did not have the time to implement many changes or impose my style. I received teaching resources from a previous member of staff and made few modifications. I adjusted the slides to make them more interesting but the model I followed was a lecture-based model where I spoke for 60-90 minutes, we took a break and the students used some of the principles in class to practice during the unassessed lab. The assessment came in the form of:
● two take-home assignments (50% of final grade)
● end of term exam (50% of final grade)
The issues I faced in class was that it was not as engaging as I was hoping it to be and this was reflected in various aspects. First off, there was a fall in attendance after a few weeks. Another symptom was that early on I got the impression that many students did not follow the syllabus every week, counting on their cramming skills to perform in the exam. Finally, the work I saw in the longer take-home assignments was not of particular high quality despite the students reporting a lot of stress around the deadline. Once again, I felt that there was a lot of frantic activity to catch-up.
Final exams are a form of assessment that is still used in Computing and having just joined I did not immediately see how counterproductive it was the use of one for a course like mine. Grades on the exam were not particularly high on average and there were lots of students that did badly. At the end of the course I felt disappointed. About 20% of the students had failed and had to resit the exam and I felt the learning objectives for the course were not met. There was a disconnect between me and the students as well as between the students themselves. I was determined to spend a significant part of the summer re-writing the module with another colleague who was going to co-teach it with me in its second iteration.
A way forward
In making improvements for the second iteration of the module we wanted to focus on these key principles:
Principle 1: Allow students time to learn at their own pace by providing all material in advance, a format known as flipped classroom.
Principle 2: Active learning reduces the gap between students in class and improves learning (Biggs & Tang, 2007).
Principle 3: Frequent small assignments promote engagement (Challis, Robinson, & Thomlinson, 2010).
Students should be given the time and space to work, think, research and come up with solutions without stress.
The first thing we decided to do was record all lectures in online videos. Instead of lecturing to 70 students for an hour or more with arguably mixed results we broke up each week’s lecture into three 20 min videos. Figure 1 shows a snapshot of me explaining color spaces using rich media. Students were given access to these videos a week before class so that they could prepare. This knowledge was then taken for granted when they came to class. This model is known as a flipped classroom and research shows that it improves the learning experience of students (Karabulut-Ilgu, Jaramillo Cherrez, & Jahren, 2018).
The next big change we implemented was to remove the monolithic exam at the end of the term. We felt that it is not the best way to assess the grasp of the learning material or reach the teaching objectives. A major part of the teaching objectives was to teach them graphics programming, something that should not be assessed in a two-hour exam. Students should be given the time and space to work, think, research and come up with solutions without stress. Instead, assessment was spread over the entire course like this:
● 8 Lab Kahoot quizzes 10%
● 8 Lab assignments 30%
● 4 Take-home quizzes 20%
● 4 Take-home assignments 40%
Shifting the marking scheme to the above was focusing on active learning as the students had to do lots of activities in order to do well in class.
When students first arrived they took a fun, often humorous Kahoot! quiz using their smartphones to indicate answers. This was intended to reward those who arrived on time and to identify points in the take home videos that were not entirely understood. Based on my impressions from the Kahoot, I would iron out any points that were misunderstood before releasing the lab assignments on the online learning platform.
The lab assignments were timed to take about two hours to complete and always had an exciting deliverable associated with them. Despite needing two hours to complete they were relatively straight forward and functioned to consolidate the students’ understanding. Working in pairs, without exception, they often achieved perfect score. All Kahoot quizzes were auto-marked and all lab assignments were designed in a way that their rubric-based marking took place at the end of the class and lasted less than a minute per pair of students, therefore eliminating the need to mark at home. This was important as it allowed us to have multiple assessments without burdening the teaching staff with long marking hours. Another positive outcome was that students got their feedback immediately and knew exactly how they were doing on the module.
The maximum points possible by showing up to class was 40%, that is, the pass grade in undergraduate teaching. This was to further encourage students to take part in the weekly class, reduce absences and make sure that they don’t fall behind. Our message to them was “Show up and engage during class and you’ll probably pass”.
Similarly, the take home assignments were designed to be fun and engaging but were significantly harder than the lab assignments. Students were expected to spend a few hours on them, working individually and were given two weeks to hand each of them in. To further assess their understanding they also had to complete an online quiz. Once again, the quiz was auto-graded by the learning platform and the take-home assignments took less than a minute each to mark.
Not having to lecture freed valuable time to help students make things, be active, talk to each other and iron out tricky points.
The results from this change were very encouraging. Students reported great satisfaction from the course as demonstrated in module evaluations. Many of them identified it as their favourite module in their short UG journey. Material from the module was often used in later projects, including in later years, showing the influence the material had on them. Not having to lecture freed valuable time to help students make things, be active, talk to each other and iron out tricky points. Teaching staff did not have to adjust the speed for all students. Students had the luxury of watching the material at their own pace in order to prepare.
Lab days were almost like a celebration. They were loud, with students running around fixing things, exchanging ideas, teaching each other and taking pride in the work they delivered. The latter was demonstrated by numerous posts on social media of their lab work (see Figure 2).
The pass rate was greatly improved. Only about 5% of students failed and those that did were students that did not engage with the class from the beginning for personal reasons. There wasn’t someone that took the course seriously and did not pass, which was very rewarding for both students and staff.
An added bonus was that the format of the class (flipped classroom + lab assignments) allowed for an easy repeat the year after. During the 2nd year running a strike was held on campus. Despite the fact I participated in it, the class was able to continue as my lab assistants could easily deliver the class without much disruption.
The class has been delivered in the revised format two years in a row with a lot of success. At the end of the second year running, I reflected on the things that would need improvement for next year:
- Everyone passed in the 2nd year running. This second year everyone passed which on one hand sounds positive, but I find somewhat worrisome. Perhaps some of the mark spreading needs to change so that there’s a “healthy” number of people that fail. I am aware of people that didn’t engage with the material yet managed to scrape by with a 40%.
- Issue with plagiarism. The homework assignments are designed to have about 70% of the points given based on a well-defined rubric, to make marking easy, and about 30% are from open-ended parts that require initiative and creativity. I became aware of many people completing the steps getting them to 70% and leaving out the bit that gets them 30%. This was an indication to me that there must be some plagiarism involved. The 70% bit is unavoidably looking very similar and it’s difficult to identify as plagiarized. Then when it comes to put their own stamp on the exercise some students avoided the 30%. A way to fix it is to shift more weight to the creative bit to reward the students that truly engage with the material.
Overall, I am very pleased with the effect the module had on the students. Because of its success I was asked to present it to the department and others have started to model parts of their module on this.
Biggs, J., & Tang, C. (2007). Teaching for quality learning at university (3rd ed.). McGraw-Hill, Society for Research into Higher Education/Open University.
Challis, N., Robinson, M., & Thomlinson, M. (2010). Coursework, what should be its nature and assessment weight? MSOR Connections, 10(2), 31-34.
Karabulut-Ilgu, A., Jaramillo Cherrez, N., & Jahren, C. T. (2018). A systematic review of research on the flipped learning method in engineering education. British Journal of Educational Technology, 49(3), 398-411.