Testing is the best teacher

In an educational world where there is a constant drive to find new and entertaining ways to support learning, it’s intriguing to see a major research project publishing its findings that, actually, traditional tests of recalled knowledge are significantly more beneficial for learning than either ‘cramming’ or the increasingly popular concept mapping.

The research, published in this month’s Science magazine and summarised in an excellent article in a recent New York Times, highlights the clear improvement in performance learning supported by assessment provides over these other approaches.  It’s also noteworthy that cramming, although markedly less effective than testing, was found to be more effective than concept mapping.

The article also highlights what is for me the most fascinating part of this research: an inverse relationship between learners’ confidence in their knowledge and the quality and quantity of what they actually recalled.  Students who repeatedly reviewed a text or who drew concept maps while consulting the text had higher levels of confidence in their knowledge of that text than those students who read the text once and then sat a test based on their recall of it.  However, in subsequent testing a week later the students who had learned the text through testing consistently performed significantly better than those who had learned by concept mapping or cramming.

One commentator suggests that this is because ‘the struggle [to recall] helps you learn, but it makes you feel like you’re not learning.’  This really does seem to get towards the core tension between making learning ‘fun’ and feel productive, and maximising what students actually learn.  Concept mapping can allow learners to illustrate in great detail what they (think they) know, but it provides no challenge to that knowledge, no way of pointing out the gaps in knowledge.  It’s a world of unknown unknowns, with the misplaced confidence it inspires demonstrating the Dunning-Kruger effect in unfortunate action.

Whatever the underlying reasons behind the findings of this research, it’s good to see the debate being opened up again and the encouragement for a re-evaluation of received wisdom about the best ways to learn.