Innovative teaching initiatives are pushing for the introduction of new tools and techniques in educational programs. Four program managers who have deployed some of these new teaching methods in their curriculum, such as a “flipped classroom” model and remote courses, were interviewed to find out the techniques’ strengths, challenges, replicability potential and possible areas of improvement.

- Prof. Christophe Demazière, Chalmers University of Technology
- Associate Prof. Jan Dufek, KTH Royal Institute of Technology
- Dr. Vladimir Radulović, Jožef Stefan Institute
- Prof. Frederic Fol Leymarie, Goldsmiths University of London

**How would you describe your innovative pedagogical method or approach?**

The pedagogical method I used is based on a flipped classroom design that I further developed and adjusted to my needs. In this set-up, some of the learning elements are offered asynchronously, so that more time is available for the synchronous sessions, during which interactions between the teacher and the students occur.

To be more precise, the asynchronous elements are made of: a textbook specifically written for the course, short videos summarizing the key points of each chapter/section of the textbook, and online quizzes focusing on conceptual understanding. The students should work with these resources before attending the synchronous sessions and they can contact/send questions to the teacher ahead of the sessions. The sessions start with a brief summary of the key concepts the students should have learned before the sessions and with addressing the questions raised by the students, either via open questions or via the online quizzes. The main part of the synchronous sessions is thereafter dedicated to activities designed to promote student learning (active learning). When possible, coding assignments using a web-based platform providing immediate feedback to the students were utilized.

The main advantage of the developed set-up is that the synchronous sessions are entirely devoted to interactions between the teacher and the students, and at supporting their learning. Active learning techniques were demonstrated to lead to significantly better student learning.

I use the following methods in my courses (not all of the methods in each course though):

• Flipped classroom approach. Students study new topics outside the classroom from materials I provide or specify. In the classroom, I briefly summarize the new topics and devote the main part of the available time to a Q/A session and other activities.

• Pre-recorded lecture topics. Students can access my voice commented slides online (YouTube) and use them when preparing for a lecture.

• Quizzes. A majority of students prepare for lectures in advance only when they are motivated, so I use graded quizzes for this reason. Graded quizzes make the flipped classroom approach effective. I give the quizzes to students during the last 10 minutes of each lecture.

• Graded and non-graded individual electronic home assignments. We have implemented a large number of assignments with solutions into the Möbius platform. Each assignment contains parameters randomly selected for each individual student so that solutions can’t be copied easily. Students can train on a number of non-graded assignments before approaching the graded assignments.

• I use other elements, such as group projects, that are effective and important but perhaps not seen as innovative today.

The Jožef Stefan Institute offers practical exercises on a research reactor to students and professionals but in-person exercises not possible due to COVID-19 induced lockdown or travel restrictions persisting after lockdown. Therefore remote, interactive exercises using off the shelf components were offered.

It is about coupling subjects (topics) that are considered abstract (and perhaps even boring for many) with concrete, visual (in this case) applications and programming exercises that give the student the experience of concrete manipulations of the topics and concepts discussed.

The traditional way of teaching maths (and many other scientific fields) is the opposite (still often today) and is done through books (books and other documents), exams/tests/exercises on paper (or the equivalent).

I make little or no use of tests/exams and instead assess progress and understanding through concrete projects (with reports, videos/demos, code and sometimes class presentations). Projects are individual and in small groups (2 or 3) each term.

**How have you got the idea to implement this new technique in this curriculum?**

The idea of flipping my classes came from the fact that in one of my courses given in a traditional onsite campus format, one of the students was remotely located. I thus livestreamed the in-class sessions for this student. I also recoded those sessions and made them available to all students. In the course evaluation for that year, the recording of the sessions was mentioned as one of the most appreciated features of the course. This was the start of a profound teaching reform, still on-going today, where the efficacy of my teaching is constantly evaluated and activities supporting student learning are designed.

We naturally try to improve our courses, and the above techniques represent an evolution of teaching. Some of them I learned about during pedagogic courses and others from colleagues.

The implementation of remote capabilities for our hands-on education activities was mainly in response to the travel restrictions in 2020 due to the Covid-19 pandemic. The JSI is regularly performing hands-on educational activities for Nuclear Engineering students of the Univ. of Ljubljana, future Krško Nuclear Power Plant operators and at the international level in collaboration with several universities, e.g. MIT, Uppsala University, University of Pavia, with in person attendance. In 2020 we needed to adapt to the circumstances and shift all our activities to an on-line format.

Together with my colleague, Prof. William Latham (who used to run his own games studio), we created the entire MSc (Computer Games Programming) and designed the various core modules.

I took charge of creating and teaching various modules, including Maths & Graphics for games (two terms), Physics in games, and more recently re-designed the AI in games module which I teach as well in the past two years.

To answer more precisely your question: maths + graphics for games is seen as core/fundamental to the development (and to some extent the design) of modern computer games. The two units (one per term) were conceived to provide the foundations needed for the programming modules.

My own background is mainly in computer vision and graphics, where it is clear that mathematics (and physics) play fundamental roles.

William and I did interview various CEOs and CTOs (and other senior people) in the games industry in the UK when we designed the various modules/courses for the MSc (back in 2007), which helped us shape and motivate the content and flow of modules, projects and other activities (seminars, game jams, etc.).

**What is the added value of this technique compared to the traditional techniques?**

The added values of the technique are as follows:

• Students in better control of their learning.

• Much more engagement from the students, both in-class and out-of-class.

• Possibility for the teacher to better tune the in-class sessions thanks to the monitoring of student learning in the out-of-class activities.

• Improved learning outcomes for the students.

• Streamlining of some of the teacher’s duties, thus allowing to focus on what matters for student learning.

• Better teacher availability to support students when they most need help.

.

The added value is not the same for all students. Some students actually prefer the traditional structure of lectures, but that may be because traditional lectures don’t require activity from students as much. For me, the outcome (in the form of achieved skills and understanding) is the significant criterion for judging the effectiveness of teaching and not only the student’s feedback. The added value is having greater flexibility during lessons, immediate feedback to student’s electronic solutions and even distribution of learning activities over the whole course period.

Readiness to perform hands-on activities in case of new lockdowns/restrictions.

Possibility of a significantly wider coverage for experimental education activities: performing hands-on activities for students located on different continents, in case of no/limited possibilities to travel in person, or for students from countries not having access to a research/training nuclear reactor.

Mixing maths and graphics makes the maths directly linked to concrete (and in this case visual, and even interactive) examples for any topic introduced (numbers and coordinate systems, vectors (arrows), matrices (2D number machines having graphical uses, in particular linear transforms in graphics (rotations, scaling, etc.)), interpolation (splines, etc.), and (much) more. At each step we have (as a teacher) at our fingertips concrete examples and coding samples and algorithms to bring to fore.

I have now taught mathematics this way to post-graduate students for 13 years and I find it much more intuitive and effective than the traditional ways (I was myself exposed to as a student). Interestingly (historically speaking) it is Descartes who linked Algebra to Geometry (made it popular) and the teachers of maths then started to favor the algebraic ways (abstract, number and equation systems) and have ignored more and more the close relation that exist with geometry (which used to be the main way of teaching/introducing concepts since the ancient Greeks). And today, it is via the same idea of Descartes that we can link back novel algebraic systems & techniques to geometry and visualisation of these concepts.

My feeling is that this way of teaching (many of the fundamental) maths concepts is exploitable in other application domains, such as : Newtonian physics, some chemistry, visual simulations in science (e.g. fluid dynamics), to name a few.

**What have been the biggest difficulties to implement it?**

• Skepticism from both students and other colleagues.

• Finding time to make course reforms.

• Administrative system for education not adapted to a different teaching approach.

• Too many different IT systems not always compatible/well integrated.

• IT systems having a too short lifespan.

• Designing engaging active learning sessions.

Some of the activities, such as the implementation of electronic assignments, are time-consuming. And, once some assignments are implemented, students quickly discover bugs in them. It’s tedious work.

The implementation required an initial investment (procurement of audio/video equipment, computers, programme licenses, etc.) and human resources (installation and configuration of equipment, testing to obtain operational feedback, corrective measures).

None really (given my own education and background: so perhaps this is the initial challenge : have/find teachers already exposed to both the algebraic and geometric way of thinking.

Good teaching material (e.g. online books) was available also by the time I initiated my own modules (in 2008). And more has emerged since.

**What are the prerequisites to implement it?**

• Perseverance!

• Time.

• Need to develop a strategy aimed at improving student learning. The strategy should be specific to one’s own course.

• Collegial support from innovators in education.

• Support from central administration.

One needs to have electronic platforms for activities that require them. We use Möbius for assignments and Canvas for quizzes and exchange of digital materials. Free access to the Internet is required for students to participate.

Available funds for required equipment and available human resources (skills in IT/audio/video, etc.) to optimally configure the systems.

Having teachers with a solid background in the relevant disciplines. Having a background in programming/computing can help too so as to speak the same language as the students who will implement projects. E.g., if teaching concepts relevant to machine learning and modern AI, it is useful to have teachers exposed to the Python** language (mainly because of the very dynamic ecosystem that exists worldwide). One needs not be an expert or practice every day, but needs not to be turned off by code examples and be fluent enough to understand (and/or learn on the way) the coding examples students will be exposed to.

** I learned my AI way back with the Lisp language .. now obsolete. I had to expose myself to Python. Thankfully, Python is much more intuitive than Lisp (at least initially).

**What would be the next steps to go one step forward?**

• Better integration of the off-site students with the on-site students, so that groups of on-site and off-site students could be formed both for the in-class and out-of-class activities. This requires designing activities allowing such mixed group activities, as well as software/hardware solutions making this possible.

• More centralized and coordinated efforts, instead of individual initiatives.

We are continuously expanding our e-learning materials, so we are improving our teaching step by step. The most important thing is to fuel the passion for physics in students and for that the traditional teacher-student interaction is essential and no hi-tech method can substitute it.

Enabling more direct interaction between students at a remote location and the experimental equipment.

Upgrading the audio/video equipment to include body-mounted cameras (e.g. Gopro), improving the audio quality and video resolution.

Initially, produce enough material to sustain the core modules where maths is introduced or needed.

Next, if such (or some of the) material needs to be built “from scratch” put together the resources to have a production team to sit along the teachers and help them produce excellent material (visuals, audio, video, interactive ideally).

Such material, if of high quality, can then be used not just in class, but also to promote the teaching/teachers, the programs themselves.

A further step would be to make such (or some) material available in various languages (French, English, German, Chinese, etc.)

Another “step” would be to review and revise/update the material every few years (say 3 or 4 or more), depending on the topics (e.g. in Machine Learning, we need to update the material more frequently than say in the fundamentals of graphics).

**How could it be replicated for other curricula?**

• Making focused “teaching materials” available to the students before the in-class sessions.

• Making short video recording of lectures, not repeating the reading the student have to do, but extracting the main features.

• Using online quizzes providing formative feedback to the students.

• Designing engaging active learning sessions.

• Mixing on-site and off-site students (flexibility required).

We use generic teaching methods that are not conditioned for specific topics, so the application of the methods in other courses is possible.

Instating remote capabilities for other curricula is very similar to our case, as it is essentially a matter of setting up a suitable audio and video connection.

One would need to seek the concrete examples that can be exposed to students for (preferably) each new mathematical concept introduced. In Games and Graphics, this is natural to do (and it is easy to find numerous examples).

It may be more challenging in some other disciplines.

For “reactor physics and safety” I would guess that some of the physics and the simulation benefits from a number of concrete examples already, that can be integrated along with the relevant maths.

For the concepts which overlap with disciplines like graphics and games, examples can be found there of course, and then adapted.

**How would you describe your innovative pedagogical method or approach?**

The pedagogical method I used is based on a flipped classroom design that I further developed and adjusted to my needs. In this set-up, some of the learning elements are offered asynchronously, so that more time is available for the synchronous sessions, during which interactions between the teacher and the students occur.

To be more precise, the asynchronous elements are made of: a textbook specifically written for the course, short videos summarizing the key points of each chapter/section of the textbook, and online quizzes focusing on conceptual understanding. The students should work with these resources before attending the synchronous sessions and they can contact/send questions to the teacher ahead of the sessions. The sessions start with a brief summary of the key concepts the students should have learned before the sessions and with addressing the questions raised by the students, either via open questions or via the online quizzes. The main part of the synchronous sessions is thereafter dedicated to activities designed to promote student learning (active learning). When possible, coding assignments using a web-based platform providing immediate feedback to the students were utilized.

The main advantage of the developed set-up is that the synchronous sessions are entirely devoted to interactions between the teacher and the students, and at supporting their learning. Active learning techniques were demonstrated to lead to significantly better student learning.

I use the following methods in my courses (not all of the methods in each course though):

• Flipped classroom approach. Students study new topics outside the classroom from materials I provide or specify. In the classroom, I briefly summarize the new topics and devote the main part of the available time to a Q/A session and other activities.

• Pre-recorded lecture topics. Students can access my voice commented slides online (YouTube) and use them when preparing for a lecture.

• Quizzes. A majority of students prepare for lectures in advance only when they are motivated, so I use graded quizzes for this reason. Graded quizzes make the flipped classroom approach effective. I give the quizzes to students during the last 10 minutes of each lecture.

• Graded and non-graded individual electronic home assignments. We have implemented a large number of assignments with solutions into the Möbius platform. Each assignment contains parameters randomly selected for each individual student so that solutions can’t be copied easily. Students can train on a number of non-graded assignments before approaching the graded assignments.

• I use other elements, such as group projects, that are effective and important but perhaps not seen as innovative today.

The Jožef Stefan Institute offers practical exercises on a research reactor to students and professionals but in-person exercises not possible due to COVID-19 induced lockdown or travel restrictions persisting after lockdown. Therefore remote, interactive exercises using off the shelf components were offered.

It is about coupling subjects (topics) that are considered abstract (and perhaps even boring for many) with concrete, visual (in this case) applications and programming exercises that give the student the experience of concrete manipulations of the topics and concepts discussed.

The traditional way of teaching maths (and many other scientific fields) is the opposite (still often today) and is done through books (books and other documents), exams/tests/exercises on paper (or the equivalent).

I make little or no use of tests/exams and instead assess progress and understanding through concrete projects (with reports, videos/demos, code and sometimes class presentations). Projects are individual and in small groups (2 or 3) each term.

**How have you got the idea to implement this new technique in this curriculum?**

The idea of flipping my classes came from the fact that in one of my courses given in a traditional onsite campus format, one of the students was remotely located. I thus livestreamed the in-class sessions for this student. I also recoded those sessions and made them available to all students. In the course evaluation for that year, the recording of the sessions was mentioned as one of the most appreciated features of the course. This was the start of a profound teaching reform, still on-going today, where the efficacy of my teaching is constantly evaluated and activities supporting student learning are designed.

We naturally try to improve our courses, and the above techniques represent an evolution of teaching. Some of them I learned about during pedagogic courses and others from colleagues.

The implementation of remote capabilities for our hands-on education activities was mainly in response to the travel restrictions in 2020 due to the Covid-19 pandemic. The JSI is regularly performing hands-on educational activities for Nuclear Engineering students of the Univ. of Ljubljana, future Krško Nuclear Power Plant operators and at the international level in collaboration with several universities, e.g. MIT, Uppsala University, University of Pavia, with in person attendance. In 2020 we needed to adapt to the circumstances and shift all our activities to an on-line format.

Together with my colleague, Prof. William Latham (who used to run his own games studio), we created the entire MSc (Computer Games Programming) and designed the various core modules.

I took charge of creating and teaching various modules, including Maths & Graphics for games (two terms), Physics in games, and more recently re-designed the AI in games module which I teach as well in the past two years.

To answer more precisely your question: maths + graphics for games is seen as core/fundamental to the development (and to some extent the design) of modern computer games. The two units (one per term) were conceived to provide the foundations needed for the programming modules.

My own background is mainly in computer vision and graphics, where it is clear that mathematics (and physics) play fundamental roles.

William and I did interview various CEOs and CTOs (and other senior people) in the games industry in the UK when we designed the various modules/courses for the MSc (back in 2007), which helped us shape and motivate the content and flow of modules, projects and other activities (seminars, game jams, etc.).

**What is the added value of this technique compared to the traditional techniques?**

The added values of the technique are as follows:

• Students in better control of their learning.

• Much more engagement from the students, both in-class and out-of-class.

• Possibility for the teacher to better tune the in-class sessions thanks to the monitoring of student learning in the out-of-class activities.

• Improved learning outcomes for the students.

• Streamlining of some of the teacher’s duties, thus allowing to focus on what matters for student learning.

• Better teacher availability to support students when they most need help.

The added value is not the same for all students. Some students actually prefer the traditional structure of lectures, but that may be because traditional lectures don’t require activity from students as much. For me, the outcome (in the form of achieved skills and understanding) is the significant criterion for judging the effectiveness of teaching and not only the student’s feedback. The added value is having greater flexibility during lessons, immediate feedback to student’s electronic solutions and even distribution of learning activities over the whole course period.

Readiness to perform hands-on activities in case of new lockdowns/restrictions.

Possibility of a significantly wider coverage for experimental education activities: performing hands-on activities for students located on different continents, in case of no/limited possibilities to travel in person, or for students from countries not having access to a research/training nuclear reactor.

Mixing maths and graphics makes the maths directly linked to concrete (and in this case visual, and even interactive) examples for any topic introduced (numbers and coordinate systems, vectors (arrows), matrices (2D number machines having graphical uses, in particular linear transforms in graphics (rotations, scaling, etc.)), interpolation (splines, etc.), and (much) more. At each step we have (as a teacher) at our fingertips concrete examples and coding samples and algorithms to bring to fore.

I have now taught mathematics this way to post-graduate students for 13 years and I find it much more intuitive and effective than the traditional ways (I was myself exposed to as a student). Interestingly (historically speaking) it is Descartes who linked Algebra to Geometry (made it popular) and the teachers of maths then started to favor the algebraic ways (abstract, number and equation systems) and have ignored more and more the close relation that exist with geometry (which used to be the main way of teaching/introducing concepts since the ancient Greeks). And today, it is via the same idea of Descartes that we can link back novel algebraic systems & techniques to geometry and visualisation of these concepts.

My feeling is that this way of teaching (many of the fundamental) maths concepts is exploitable in other application domains, such as : Newtonian physics, some chemistry, visual simulations in science (e.g. fluid dynamics), to name a few.

**What have been the biggest difficulties to implement it?**

• Skepticism from both students and other colleagues.

• Finding time to make course reforms.

• Administrative system for education not adapted to a different teaching approach.

• Too many different IT systems not always compatible/well integrated.

• IT systems having a too short lifespan.

• Designing engaging active learning sessions.

Some of the activities, such as the implementation of electronic assignments, are time-consuming. And, once some assignments are implemented, students quickly discover bugs in them. It’s tedious work.

The implementation required an initial investment (procurement of audio/video equipment, computers, programme licenses, etc.) and human resources (installation and configuration of equipment, testing to obtain operational feedback, corrective measures).

None really (given my own education and background: so perhaps this is the initial challenge : have/find teachers already exposed to both the algebraic and geometric way of thinking.

Good teaching material (e.g. online books) was available also by the time I initiated my own modules (in 2008). And more has emerged since.

**What are the prerequisites to implement it?**

• Perseverance!

• Time.

• Need to develop a strategy aimed at improving student learning. The strategy should be specific to one’s own course.

• Collegial support from innovators in education.

• Support from central administration.

One needs to have electronic platforms for activities that require them. We use Möbius for assignments and Canvas for quizzes and exchange of digital materials. Free access to the Internet is required for students to participate.

Available funds for required equipment and available human resources (skills in IT/audio/video, etc.) to optimally configure the systems.

Having teachers with a solid background in the relevant disciplines. Having a background in programming/computing can help too so as to speak the same language as the students who will implement projects. E.g., if teaching concepts relevant to machine learning and modern AI, it is useful to have teachers exposed to the Python** language (mainly because of the very dynamic ecosystem that exists worldwide). One needs not be an expert or practice every day, but needs not to be turned off by code examples and be fluent enough to understand (and/or learn on the way) the coding examples students will be exposed to.

** I learned my AI way back with the Lisp language .. now obsolete. I had to expose myself to Python. Thankfully, Python is much more intuitive than Lisp (at least initially).

**What would be the next steps to go one step forward?**

• Better integration of the off-site students with the on-site students, so that groups of on-site and off-site students could be formed both for the in-class and out-of-class activities. This requires designing activities allowing such mixed group activities, as well as software/hardware solutions making this possible.

• More centralized and coordinated efforts, instead of individual initiatives.

We are continuously expanding our e-learning materials, so we are improving our teaching step by step. The most important thing is to fuel the passion for physics in students and for that the traditional teacher-student interaction is essential and no hi-tech method can substitute it.

Enabling more direct interaction between students at a remote location and the experimental equipment.

Upgrading the audio/video equipment to include body-mounted cameras (e.g. Gopro), improving the audio quality and video resolution.

Initially, produce enough material to sustain the core modules where maths is introduced or needed.

Next, if such (or some of the) material needs to be built “from scratch” put together the resources to have a production team to sit along the teachers and help them produce excellent material (visuals, audio, video, interactive ideally).

Such material, if of high quality, can then be used not just in class, but also to promote the teaching/teachers, the programs themselves.

A further step would be to make such (or some) material available in various languages (French, English, German, Chinese, etc.)

Another “step” would be to review and revise/update the material every few years (say 3 or 4 or more), depending on the topics (e.g. in Machine Learning, we need to update the material more frequently than say in the fundamentals of graphics).

**How could it be replicated for other curricula?**

• Making focused “teaching materials” available to the students before the in-class sessions.

• Making short video recording of lectures, not repeating the reading the student have to do, but extracting the main features.

• Using online quizzes providing formative feedback to the students.

• Designing engaging active learning sessions.

• Mixing on-site and off-site students (flexibility required).

We use generic teaching methods that are not conditioned for specific topics, so the application of the methods in other courses is possible.

Instating remote capabilities for other curricula is very similar to our case, as it is essentially a matter of setting up a suitable audio and video connection.

One would need to seek the concrete examples that can be exposed to students for (preferably) each new mathematical concept introduced. In Games and Graphics, this is natural to do (and it is easy to find numerous examples).

It may be more challenging in some other disciplines.

For “reactor physics and safety” I would guess that some of the physics and the simulation benefits from a number of concrete examples already, that can be integrated along with the relevant maths.

For the concepts which overlap with disciplines like graphics and games, examples can be found there of course, and then adapted.

**Follow us to find out more!**

[…] Interviews with 4 program managers who have deployed some new teaching methods in their curriculum, can be read here. […]