The Joy of Confirmation Bias

I wish I had read The Black Swan sooner. It is an engaging if sometimes challenging read. Taleb is a somewhat unconventional writer and he does not always explain complex concepts with the layman in mind. (I was initially floundering on the use of fractals in probability calculations…) But its basic message is worth the effort, and confirmation bias alone sees to it that it is satisfying to read something that justifies one’s own prior thoughts!

I show below a table reproduced from the later stages of the book. In it Taleb summarises contrasts in ways which different paradigms approach uncertainty. A Black Swan is a high-impact, low-frequency event that nonetheless has a massive effect; for instance, Taleb claims that 50% of the value of American stocks has been added by just ten tumultuous days in the last 50 years; similarly, just one blockbuster can completely turn a publisher’s fortunes around. Getting lucky enough to ride such a wave can be transformative, as can the negative impact of unforeseen catastrophes.

Taleb.jpg

The salient point for teaching, I think, is that there are aspects of life where grand master-plans and scientific predictions simply do not wash. Taleb points out that the social sciences and human behaviour in general are amongst them. There is no point in organising such phenomena around the diktats of elegant theories, or statistical predictions: the range of possible human responses to educational acts is just too great to call, and it does not regress to the mean either – something which most models assume. It is the educational equivalent of the Butterfly Effect.

One might conceive of a small event in a pupil’s schooling, that unbeknown to the teacher ‘lights a bulb in the mind’, and eventually causes that pupil to become the next Bill Gates or Stephen Hawking. I wonder what effect size John Hattie would ascribe to that… But to a lesser extent, the same is true about the actual effects of any educational (or other) interaction; this is why trying to identify and then depend on known ‘outcomes’ is pointless. (One might also consider the effect on the ‘total career impact’  of a teacher, of the act that produces the next Einstein!).

When I was writing my own book, I felt slightly uneasy that I could not substantiate sufficiently my claim that educational interactions are ‘irreducible events whose outcomes are only known to those involved’. For all that it felt right, justifying it was more difficult. But here, at last, is something that I think comes close. If it is indeed true that some effects in this world are governed by a fractal model of probability rather than a Bell Curve, then it is entirely possible for the effects of a teacher on a pupil to be unknowable – an educational Black Swan.

This does not mean that extreme events will occur frequently, but equally it does mean that they are eminently possible, and that we should not seek to build theories without allowing for them. Likewise, for every pupil who ‘works hard and succeeds’, how many (often unseen) pupils work hard but don’t? Or don’t work hard, but still do? It’s a deeply misleading relationship to suggest to people.

But given that they are unknowable in both nature and occurrence, we cannot therefore build accurate models of our effect. Resorting to those we have on the grounds that they are better than nothing does not make them any more accurate.

Taleb suggests that the only sensible response to this is Sceptical Empiricism, in other words a mode of working that is deductive: it works from reality to concept, rather than the opposite, always remembering that the past is not a good predictor of the future. The chart summarises the differences between this approach and the more conventional one, which does indeed look more like the present-day character of some education systems.

I was delighted to note that Taleb describes this approach as being ‘sophisticated craft’ rather than ‘poor science’ – which is indeed exactly how I would characterise my view of teaching vis à vis the established one, and what I proposed in the conclusion of my own book. And as he says, it is better to be broadly right than precisely wrong.

Which is what the current models being used in education are.

Advertisements

One thought on “The Joy of Confirmation Bias

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s