This paper presents three interventions intended to increase the conceptual understanding of students in the Computer Systems course at University of Copenhagen Department of Computer Science. This is a bachelors level course introducing to students the topics of computer architecture, networking, and operating systems. The three interventions in the course are: switching lectures to be more activity based, releasing recordings of the old lectures, and asking students more conceptual questions through their assignments. The intention is that the recordings will free up contact time, so that activities and assignment questions can focus less on the raw facts of computer science and more on the key underlying topics.

Results are promising, with students demonstrating sophisticated answers according to the SOLO taxonomy in several assignment questions. Most students were capable of hitting the higher Relational or Extended Abstract tiers in at least some questions, while other questions did not provide a single high-quality answer. This is taken to be less a critique of the students, but that material needs to be deliberately designed as if students do not need to demonstrate the higher tiers of knowledge to answer a question, they will not do so. Grades were not significantly different from previous years, though the focus of the material in both assignments and exams has shifted from more fact based questions to those focussing on conceptual understanding.

A notable drop in attendance was observed and so the second half of this paper investigates what specifically caused this and evaluating how much of a problem it is. This was primarily done through surveys and conversations with students with their primary reasoning being the availability of online resources including the provided recordings, but also LLMs. As final grades did not drop despite the lower attendance a brief investigation was carried out correlating participation with final outcome and it is shown that there is at least a relationship between the two, with increased participation correlating to higher student outcome though a causal relationship is not claimed.

Finally, the paper makes recommendations about what lessons could be learned from this experience. The key one is that simply having access to an experienced teacher to ask questions of is not a sufficient motivation on its own, and that students regularly feel they have enough resources available to them to learn on their own. A group and/or problem based approach to class based activities is suggested as a way of levering a social obligation to participate.

Keywords: Classification, SOLO taxonomy, Flipped Learning.