For our group work this time round, we prepared a mock syllabus for a course on Online Collaborative Writing, peppered with roll-over points describing the thinking behind our choices, using a tool called Thinglink.  This helped to concentrate our thoughts on underlying design issues, which have been laid out clearly by my colleague Sebastian Schwede.

In this blogpost, I want to take a step back from the individual course, and think about what it is we are trying to teach in a more global sense.  I teach physics, and so we are trying to impart some facts and information, but there is far more out there than can be crammed into an undergraduate course, so we engage in identifying what we think the most important bits are, and, more importantly, in training students to think “like physicists”.

Physics World (the magazine of the UK Institute of Physics) runs a monthly column called ‘Once a physicist’, interviewing people who studied physics at some point, and then moved into careers in areas that or more (e.g. sales engineer for a scientific instruments company) or less (e.g. opera singer) related.  One thing that is often mentioned in these articles is the transferable skill of being able to develop an approach to solve a problem, or check for feasibility using approximations. The archetypal problem for this latter case is “How many piano tuners are there in Chicago?”  The version I have used in my own small group teaching is usually what is the mass of an aeroplane.

Of course, it is sometimes (but not always) trivial to Google the answers to these questions.  This does not necessarily help with exploiting or understanding the answers.  For example, in the case of the aeroplane, the follow-up question would be: why do the air hostesses sometimes insist that you shouldn’t change seats on a half-full flight?

So, how do we learn (and teach) things like this?  What kind of learning are we aiming for?  There is a a huge amount of literature on this, generally arguing for deep learning as opposed to surface learning – here following the definitions of Marton and Säljö, later expanded by Biggs in his book Teaching for Quality Learning at University.

I now want to illustrate this by using several examples that may seem irrelevant.  Consider juggling.

This is a learned skill.  I learned how to juggle three balls over a summer by starting with two against a wall, then three against the wall, then three freestanding.  With time and practice, you adapt to the ball trajectories in advance, and can cope with unforeseen obstacles (a distracting friend, a gust of wind, etc).  Or consider snooker.  This can easily be visualised as Newtonian mechanics in action.  Understanding the physics can help (why should I hit the white ball there?) but being a good snooker player requires a lot of practice at the table, until your responses are built on instinct.

Or consider landing a plane.  A pilot has to study a lot of theory, but this is combined with a lot of practice so that, when there is suddenly a strong cross-wind at the landing strip, the pilot can build on their lived experience to respond appropriately.

All of these examples can be nicely described by the very clear work of Daniel Kahnemann and Amos Tversky.  I highly recommend Kahnemann’s excellent book Thinking Fast and Slow.  This assigns two systems to the brain.  System 1 is “fast, automatic, frequent, emotional, stereotypic, unconscious”, and System 2 is “slow, effortful, infrequent, logical, calculating, conscious”.

In all of the examples above, the aim is to push as much processing into System 1 as possible.  If you are using System 2 to juggle, you might understand what is going on, but the balls end up all over the floor.

For physics, the aim is not dissimilar.  By means of extensive practice and consideration we want to incorporate a ‘physics intuition’ into the brain.  I want my students to be able to look at the graph they have made from their experimental results and immediately realise that there has been a problem with the error calculation.  But, it takes time and effort.

I have some experience at trying to do this in offline courses, but how do I do this in online or blended courses?  Well, for certain algorithmic questions, it is possible to generate a whole stack of them for the students to plough through, receiving immediate feedback as they work.  There are many open educational resources, providing high quality illustrations, animations, videos and notes.  Eventually, however, consuming such resources is insufficient.

Offline, the peer-to-peer component is vital for cementing learning.  Online, this can be done with discussion boards and other fora for online engagement, but it is harder.  Nonetheless, there must be opportunities (obligations?) to engage with others in some way to jointly construct knowledge.  Obviously this can be done – here are some student reviews from the Open University, UK, for an advanced course on quantum mechanics.  One particular comment I want to highlight is the following:

This is simply the most fascinating course I have done with the OU to date. One where you make a key transition in your study of physics: from relying on your physical intuition to inform the maths; to relying on the mathematics to tell you what your intuition should be!

This is getting right at the question I wanted to address with this post, the retraining of your intuition!  Now, I just need to figure out how I can implement it myself!