One piece of research that Alter brings up in Drunk Tank Pink that was in the news a few years ago is how hard to read fonts increase success in problem solving. He specifically points in his book to research that gives subjects questions like:
A bat and a ball cost $1.10 in total. The bat costs $1 more than the ball. How much does the ball cost?
For half of the subjects, the font is rendered clearly, easy to read. For the other half, the font is lower contrast, harder to read–a condition called disfluency. Subjects in the latter condition are more successful at giving the right answer because, argues Alter, the harder to read text puts the brain in a state of alert–it wakes the brain up, if you will. The brain devotes more resources to solving the problem and is more skeptical of solutions.
Well, shouldn’t we render all training materials with some level of disfluency, then?
Maybe. But there’s more to think about. If disfluency puts the brain in an optimal state for problem solving, is there another way to get there, one perhaps that has less potential to negatively impact learners (say, learners with visual acuity issues or dyslexia–or even just annoying learners at some level if something is not easy to read)? The studies behind this were in a research setting, where subjects aren’t necessarily motivated to learn, and not in classrooms, where presumably learners have some stake in their learning. If a well-designed class is already putting participants in the right frame of mind, would disfluency just get in way?
Another critical piece of this is cognitive load. The premise in the experiments above is that subjects were capable of solving the given problem with the right insight. The problem is not hard, it’s just tricky. But what if the content is genuinely challenging, taking all available cognitive resources to apply skills being learned at the same time? At that point, does disfluency push the problem solver into cognitive overload, inhibiting problem solving?
Yue, Castel, and Bjork explored disfluency in 2012. Of particular interest is how the researchers frame the issue in terms of cognitive load, and in terms of establishing the conditions where disfluency is desirable and when it is not, which is the right direction if we want to start exploring the possibility of inserting intentional disfluency into the instructional design toolkit.