Till innehåll på sidan

Lucie Delemotte, Biophysics

Publicerad 2020-10-29
Lucie Delemotte

SI1410 Basic modelling in biotechnology is a mandatory course given in P1 of the third year of the degree program in biotechnology. In 2020, it has been given as a full online course to 75 registered students.

The aim of this class is to teach students to model and analyse dynamical systems of biotechnological interest. The course has two main parts that are designed to work together to promote learning: a pen-and-paper one and a computer-based one.

Prior to 2020, the pedagogic activities were organised in six modules, each module built around a theme lasting a week. Each module had suggested reading, one or two lectures, a workshop activity where students practiced problems in groups (of 4 to 6) with access to a teacher to ask questions, an exercise session where these problems were solved at the blackboard and a computer lab session in half-class where computational problems were solved in pairs. All study material (lecture slides and additional notes, exercise corrections were provided on canvas)

The students’ mastering of the intended learning outcomes was examined through two graded lab reports (individual, even if the labs were performed in pairs) and a in-class written exam, where pen-and-paper exercises and on-paper “programming” exercises were given.

2020 and the required switch to a full on-line format provided an opportunity to try out new pedagogical activities. Overall, the changes introduced led to considerably more work for the teachers, but appear to have increased the students' proficiency in the skills related to the in learning outcomes. I am very grateful to my teaching assistants Koushik Chowdhury and Michele Pellegrino, who put in much effort redesigning some of the teaching activities, and without whom this transition would have been much more painful! I also found ideas and inspiration from our department teachers’ meetings and KTH-wide lunch n' learn seminars.

The main changes we introduced were the following: the lectures were pre-recorded, as shorter (10-30 minutes) segments corresponding to well-contained topics; lecture hours were replaced by voluntary zoom-based Q&A with teachers or TA; weekly quizzes were introduced mid-module to test the students’ progress and provide an incentive to study videos and suggested reading material prior to the workshop and exercise session; help during workshops was available online; exercise sessions were conducted online, replacing the blackboard by the use of the Notability app on iPad; computer labs were supervised on zoom, with student pairs placed in breakout rooms.

Since this class has been given 4 times in person, we had a stable structure and a lot of material already prepared. However, the new activities took substantial time to organise and plan. We started recording lecture material, preparing quiz questions and selecting exercises from our exercise bank at the beginning of August, dedicating a half/week to each module. By the beginning of the class in P1, most of the material was ready, but we had not completely settled on an examination form, beyond keeping the computer lab reports from previous class editions.

By the middle of the period, we settled on having a remote home exam, not monitored through zoom, complemented by a short oral exam designed to verify students’ identity and mainly to discourage students from getting outside help on the home exam.

The home exam was to have two parts, a pen-and-paper part, and a computer exercise part and access to all course material was allowed. To avoid time pressure, I scheduled the home exam from 8am to 5pm, with additional time for the funka students who needed it. It was scheduled on the Monday of the exam week, and oral exams were scheduled for Thursday/Friday. This meant the exams had to be corrected and graded on Tuesday/Wednesday.

Instead of having a mid-term exam that provided additional points prior years, we designed a mid-term practice home exam, to be carried out during the blank week between modules 3 and 4. We indeed felt that providing grading and feedback for a real midterm in a timely manner would not be possible. During that week I also organised a zoom meeting between myself and the student groups. I requested they have video on as much as possible. I used this to create a feeling of belonging, to get to know the students a bit more, and to make them comfortable with me before we met in the oral exam. It was refreshing to meet with the students, and made me realise how much I had missed the face time.

Having learned half of the exam was going to be on the computational aspects, many of the students were worried. This is a program where the students do not have a lot of programming experience, and since previous class editions only had a written exam, they had not dedicated as much effort for the computational part of the course so far. Having these face-to-face meetings was a good occasion to talk with them about this and give them guidance on how best to learn this new skill, which was largely perceived as important, but challenging to acquire.

To avoid cooperation between students, we decided to design several exam texts. We settled on 7 different exams, to avoid giving the same exam to members of the same workshop groups while keeping the workload of designing exams to a manageable level. I was worried about the difficulty level of the exams to be unequal. To mitigate this, we first did our best to take into account this concern while writing exam texts. We also later adjusted the grading scheme such that the average and the min/max grade for each of the seven exams were not too different. A speaker in a lunch n’learn seminar also eased my fears by pointing out that we also can't guarantee exams are as easy from one year to the other and that students also perceive different tasks to be easy or difficult.

I actually found grading more enjoyable with this system, because it appeared less repetitive and exhausting to grade ~10 exams of the same kind as opposed to ~70 in a regular exam setting. Being able to grade PDFs directly was also much easier than grading on paper.

For the oral exams, I decided to hold 10-minute exams, with a 10 minute break every hour, on top of a 1h lunch break. I had students sign-up in an excel sheet beforehand and asked them to check their internet connection, video and audio quality prior to connecting. They were asked to log in 5 minutes in advance and were placed in a virtual waiting room until they were called in. They had to show their ID to begin the discussion. The oral exam was designed to check that they had not received external help during the written exam, and was quite successful for that purpose. However, some students expected to be able to significantly alter their written exam answers during the oral exam. I reasoned this would not be fair to all students, because several additional days were available between the deadline to hand-in the written exam and the oral one. This led to disappointment for several students and additional clarification prior to the exam will need to be corrected if we keep a similar system going forward.

Overall, the changes we made led the students to be much more proficient on the computational aspects. I don't perceive any major difference on their proficiency on the pen-and paper side. The grade distribution turns out to be quite similar to previous years, even though the computational ILOs are better fulfilled. That is due to the changes in the examination form. The course evaluation is ongoing so I don’t yet have a fine understanding of student’s perceptions of the course, but the overall feedback during oral exams was positive.

I look forward to seeing the results of the course evaluation to make decisions on which aspects of this course design to keep for next year. Indeed, even if the hope is that we are able to have in-person teaching in 2021, I think keeping a flipped-classroom approach with quizzes and recorded lecture items may be a successful approach. For the examination, I have yet to decide. The 2020 approach seems to lead to better learning, but was a lot of work for the pedagogic team. Being able to recycle some of the material we designed this year may lead me to decide to have this examination form again, but I’ll have to recover from this intense effort before being able to make this decision!