Back in 2011, there was a flurry of excitement when a group of researchers found that students recalled more information if it was presented in a hard to read font, such as something fancy in italics, than if it was presented in a traditional, easy to read font like Times New Roman.
The finding fed interest in the concept, developed by Robert Bjork and colleagues, of ‘desirable difficulties’. In short, the idea is that by having to engage more cognitive processes to read the font, we learn the material better.
This view is in apparent conflict with findings from cognitive load theory, findings that generally suggest that we should reduce any unnecessary cognitive load.
Since 2011, there have been numerous attempts to replicate this effect. Now, a meta-analysis has been published that attempts to draw this evidence together. The findings are mixed, with no overall effect on basic recall or transfer for ‘perceptual disfluency’ as it is known to researchers. However, it does pretty reliably increase learning time and reduce learners’ perceptions of their own learning.
It is tempting to leave the issue there as an effect that has failed replication, but I think it might be worth introducing another idea from cognitive load theory, that of element interactivity.
Element interactivity essentially describes the complexity of a learning task. Cognitive load theory effects have been demonstrated reliably when element interactivity is relatively high, such as when students are learning to solve algebra problems. On the other hand, desirable difficulties effects have mainly been demonstrated on tasks such as learning the names of state capital cities, something that would be relatively low in element interactivity / complexity. It would be interesting to see how effects for perceptual disfluency map against the complexity of the learning task.