06 September 2015

In just over a week’s time we will run our fourth Sheffield Gaussian process summer school. Teaching on summer schools is a great pleasure, and one reason for this is that you are not teaching to a specific measurable.

If you ever watch a children’s movie, a good children’s movie, you find it works on several levels. There are often cultural references in the film that entertain adults: very often they use famous actors for voice characterization, but these actors are unknown to the children. Children and adults perceive different things in the movie because it gives them different things.

Knowing the right way to teach an advanced concept is very hard. It is not a mechanistic process. You don’t know exactly what is in the mind of your students when they start, what their background is and what will inspire them. When you write an exam you are tasked with testing exactly what is in their mind at the end. On a university course an exam is needed for accountability. It is not part of the process of learning, it is merely a stamp of authenticity demanded by society to quantify the capabilities of the university and the student.

Teaching to such goals significantly inhibits the approaches you can take to teaching. You aren’t able to speak on different levels, and you aren’t able to incorporate a diversity of different backgrounds.

Of course such accountability is vital for many reasons, but it occurred to me the other day that the formal process of measurement significantly affects the nature of a taught module. When measuring outcome, it is not possible to set individualised exams that test the outcome for each student. If a module does work on multiple levels, it still can’t be assessed that way. You will have to define how the module will be assessed, i.e. did the monkey have a funny voice? It seems that the process of measurement therefore necessarily effects the state of the system.

The great pleasure of the Gaussian process summer school (and summer schools in general) is that we are not teaching to an exam. We are not accountable in that way. Our success is based on something a less tangible. Our accountability is the success and reputation of the school.

Being released from defining specific learning outcomes means we can teach to inspire, develop material that (hopefully) works on multiple levels. This is a lot more pleasant teaching experience than to a specified set syllabus. Of course, whether it is any more effective overall is difficult to measure.

Of course this isn’t so much a quantum theory, but it’s an idea that has an aspect of quantum mechanics in it. We will always be uncertain as to the actual quality of the underlying course, and attempting to measure it will affect it.



blog comments powered by Disqus