Desirable Difficulties

What it is

In their 2020 commentary closing a journal special issue on the idea, Robert Bjork and Elizabeth Bjork define desirable difficulties as challenges that trigger encoding and retrieval processes which support learning, comprehension and remembering. The word desirable carries a condition. A difficulty is desirable only if the learner has the background knowledge or skill to meet it. Without that, the same hurdle is simply a difficulty, and it does harm.

The framework starts from a distinction the Bjorks made in their earlier work between storage strength and retrieval strength. Conditions that speed up apparent performance during practice often fail to support retention and transfer later. Conditions that slow apparent learning often work better. The commentary says spacing and variation reduce retrieval strength, which in their theory can enhance storage.

The commentary lists the forms the idea has taken. Spacing practice out over time improves long-term retention, and interleaving is described as a form of spacing. Retrieval triggered by testing can strengthen later recall, reduce intrusions from incorrect information and show learners what they have not yet learned. Generation, producing knowledge from memory rather than being shown it, is generally desirable, provided the learner has the prior learning to succeed. Pretesting, attempting answers before instruction, and learning from errors appear as further candidates, and the authors note that errors are hard to use well in practice. Adaptive schedules that tailor difficulty to a learner's past success are offered as good examples of difficulty that is challenging but solvable.

Every item on that list changes what the learner has to do. None of them changes how the page looks.

In effect

Research: the font that was built to be hard

Sans Forgetica is a typeface designed to be hard to read, on the theory that the difficulty would help memory. Taylor, Sanson, Burnell, Wade and Garry (2020) compared it with Arial across four experiments, reported in the journal Memory under a title that states their answer: disfluent difficulties are not desirable difficulties. Readers found Sans Forgetica harder to read, and it did not improve their memory. Sometimes their recall was worse. The experience of difficulty was real and the benefit was not.

The earlier claim for fonts in education came from Diemand-Yauman, Oppenheimer and Vaughan (2011, Cognition). According to a summary hosted at Princeton, 28 volunteers learned material about extraterrestrials and scored 86.5 per cent with a harder font against 72.8 per cent with an easier one, and 222 high-school students in Chesterland, Ohio, did better on regular classroom tests when their material was set in harder fonts. The authors warned that a font could become hard enough to block learning. In 2016 Bjork and Yue's commentary on a set of six studies in Metacognition and Learning reported that they generally failed to find the benefits the 2011 paper described.

Everyday life: where the difficulty sits

Sirota, Theodoropoulou and Juanchich (2021) tested hard-to-read fonts on reasoning problems in 310 participants and found no improvement in solution rates, whatever people's numeracy. In a second experiment with 311 participants, simply instructing people to pay attention to and reflect on the verbal problems raised solution rates substantially. That is the pattern the Bjorks' definition predicts. The difficulty that helped was in the task, where the person had to think, and not in the typeface.

What it does not say

It does not say that anything hard is good for learning. The 2020 commentary is explicit that a difficulty becomes undesirable when the learner lacks the background to meet it, and that conditions which slow apparent learning are useful only where they raise retention and transfer.

It does not say that hard-to-read fonts help. The font studies above found no memory benefit from Sans Forgetica in four experiments and generally no learning benefit in the 2016 set of six studies, and the reasoning claim failed in a pooled analysis of the original experiment with 16 replication attempts. The concept and the font application are separate, and the first can stand while the second falls.

It does not say that learners will like it or notice that it works. The commentary reports that learners often feel they are learning less under desirable difficulties, and that massed practice can create an illusion of mastery. The authors write that they once assumed showing learners the research would be enough, and now see motivation as a major obstacle.

It does not say the applied evidence is uniform. The commentary discusses the difficulties of putting the idea into practice and notes that some groups, such as children, have been studied less for some of the techniques.

---