Interpreting Signals in Tertiary Education
When I began my postgrad adventures last year, I had this dreadful feeling that I was not meant to be here. I distinctly remember one of my classmates scoffing at my thesis idea and one of my lecturers telling me (the most dreaded) “that’s not very interesting”. Added to that: I had failed economics in my first year, had zero maths background and performed second-to-worst in the first test of our maths Boot Camp at the begining of the year.
It was not a promising start.
As the year progressed, my feelings of innadequacy grew but for one thing: there was a small handful of lecturers that seemed to have an irrational and illgical amount of faith in me. One of them was sweet and savvy and young and took time out of her weekend to help me with an empirical project; one of them always took the time to make sure that I was alright when I looked tired and over-worked; one asked me if I was ok once and I burst into tears in his office and he told me that I was better at this than I thought I was.
By far the most important was my supervisor. He told me that grades were no indication of ability and that everyone knew that. He spent plenty of time explaining that your work should speak for itself; that your goal should be to learn and that learning often meant making mistakes. He said that good work, good ideas and good questions were far more important than getting a distinction in a maths course.
He spent an equal amount of time telling me why these courses were important anyway. And I worked hard at all of them. I never did well but I know that I have learned a lot. And my “imposter syndrome” has changed flavour: it’s become the type that has some hope that I may actually become a talented economist but tainted with some fear and a whole lot of risk (hope, as it turns out, results in fear and implies risk. So much so that I would hazard to say that “risk-loving” is simply another way of saying “excessively hopeful”)*.
I still don’t do well on my tests.
Recently, I had a fascinating conversation with a lecturer. He made one, simple statement which, at the time, came as a huge relief. Later, it became a point of significant concern:
“If you produce good work, nobody is going to think you’re a bad economist just because you got 50% for your maths course.”
Before someone jumps down my throat for this (potentially) controversial statement: note the qualifier. If you produce good work.
From most of what I have witnessed, this is an accurate statement. And, in many ways, I am relieved because I know that I am capable of producing good work. But the economist in me can’t help but be concerned…
Courses are there to teach you skills that enable you to do the work required of an economist. Grades are meant to act as a signal of how well you’ve internalised these skills. And much of my future is dependent on how well I signal my ability — wherever it is that I end up doing my PhD will be determined by where it is that someone thinks I have talent as an economist.
In general, my department thinks that I possess this talent (otherwise they would not be suggesting that I do my PhD). But my grades are not accurate signals of this. And, strangely, many people seem to be aware of that. Economics predicts that unreliable or inaccurate signals won’t be used…
But we still use grades as signals. Is it because they are all we really have? If so, why have we not found others? Or why have we not found ways to improve the reliability of grades as signals? Fundamentally, the statement made by my lecturer indicates that grades are not effectively or efficiently accomplishing their required tasks and, as educators and economists, their continued use at a postgraduate level is an example of misbehaving that I can’t quite explain.
Actually, I lie. I can attempt to explain it… Either the tertiary education system is being heavily influenced by some status quo bias (or naturalistic fallacy) or I am exhibiting the Dunning-Kruger Effect. Likely, it is the latter (in which case, is it still the Dunning-Kruger Effect, if I am aware of it?) but, given the goals and desired outcomes of education, it is worth considering the possibility that it is the former. In which case, we ought to be asking:
(1) Should we be using grades as signals at all?
(2) If yes, how do we improve their reliability?
(3) If no, what could we use instead?
*Keep an eye out in the next few weeks: my next blog post will (kind of) be a sequel to this one. Specifically: on learning and talent.
