[2018-01-31 Wed]

I have decided that next year I am going to try lecture-flipping my topology module. Lecture-flipping is the practice whereby the lecturer prepares material for the students to view/read ahead of time, and the lecture is spent getting the students to explore the material in more depth and consolidate their understanding. I have decided to document my experience in case it's useful for other people thinking about flipping (specifically in the context of maths lectures).

In this blog post I want to review:


Here are some very explicit learning outcomes I would expect from a student coming out of a four-year mathematics degree:

By the end of this degree, you should be able to:

  1. quickly and correctly evaluate which parts of a proof are easy or routine and which require more thought or new ideas.

  2. identify gaps in mathematical arguments (your own and those of others) and evaluate whether they are easily fixed or not.

  3. extract and summarise the key new idea or construction at the heart of a proof.

  4. devise detailed, step-by-step strategies for proving previously unseen results.

Of course, we hope that all students will pick these skills up by osmosis in analysis courses, but think about it from a student's perspective:

Whereabouts in all this do they get the chance to achieve any of the learning outcomes listed before? If it happens, it happens by accident.

What the student does

There is a persuasive line of reasoning in the education literature (for example in Bigg's article "What the student does") that goes as follows.

Of course, we expect students to do their problem sheets religiously and to develop mathematical skills as they do so, but often they are working problems alone with very little idea where to get started, and all they can do afterwards is to learn by heart the model answers they are told as feedback.

Lecture-flipping puts the emphasis on students doing maths during lecture time. Maybe they can work through some examples in groups and ask for guidance when they become stuck. Maybe they can flesh out a sketch-proof, or critique one another's written proofs. They can receive immediate feedback to help them, they can start to evaluate their own work and that of their peers.

What does it mean in practice?

First, it means that much of the standard lecture content must be delivered in another format: either in books/lecture notes or online videos. Indeed, this aspect is what most people think of as lecture-flipping.

More importantly, it means that you have to think of something else to do in lecture time. You have to design meaningful learning activities for the students to do to help them to achieve the learning outcomes you want.

I have already written a comprehensive set of lecture notes for the module I'm planning to flip. My plan is the following:

  1. Restructure and divide the notes into 10-20 minute chunks and decide which part of the notes I should just tell them (in a video) and which parts they should be able to figure out for themselves (maybe in groups in class).

  2. For each chunk, produce either:

    1. a video, along with a set of comprehension questions and a set of notes for them to read alongside the video. The comprehension questions should be relatively quick and easy (the kind of things you should be able to do instantaneously/mechanically if you understood the video). They might involve
      • filling in details of a proof from the video,
      • correcting a slightly incorrect statement that was made in the video,
      • doing a slightly longer computation similar to one that was covered in the video.

    2. a lesson-plan for a face-to-face session in which we cover the content, for example:
      • 0:00-0:10 At the beginning of the class, state the result that the deck group is isomorphic to the quotient \(N_H/H\) where \(N_H\) is the normaliser of \(H\) and \(H\) is the subgroup associated to the quotient space. Get the class to remind you of the definitions of all terms involved.
      • 0:10-0:25 In pairs, work out what this means for some explicit covering spaces (need to prepare handouts: one covering space for each pair).
      • 0:25-0:30 Share your conclusions with another pair.
      • 0:30-0:40 As a whole class, formulate a strategy for proving the theorem (lead them to the idea that if \(\beta\in N_H\) then there is a deck transformation \(F_\beta\) taking \(y\) to \(\sigma_\beta(y)\) and that you need to show this assignment \(\beta\mapsto F_\beta\) is (a) a homomorphism, (b) surjective, (c) injective).
      • 0:40-0:55 In pairs, try to implement that strategy. (Homework: complete the proof. Next session, someone will present their solution).

Of course, you end up covering less content in class this way, but as I tried to emphasise earlier, the kind of skills they are developing are perhaps more important than just learning the content. For example, they are learning how to approach proofs for themselves, and how to talk constructively about maths with, and explain stuff to, their peers.

What have I already done?

The last week of term I was not in UCL and wasn't able to give my lectures in person. Having already settled upon flipping next time around, I decided to record the week's lectures as videos and set some learning activities as homework. This doesn't count as flipped learning, because it wasn't backed up with face-to-face contact time. Nonetheless, I learned a lot from doing it. Here you can see the online learning materials I provided.

Comments, corrections and contributions are very welcome; please drop me an email at j.d.evans at lancaster.ac.uk if you have something to share.

CC-BY-SA 4.0 Jonny Evans.