One of my main research questions is how does learning work, and what can we do to make memories stick? The other is, how do people think they learn best.

Learning and Remembering
There’s no point in learning something if you can’t remember it later. The real goal of education is long-term retrieval. The spacing effect and the testing effect, which I focus on, make learning more difficult in the short-term but enhance long-term retention–in other words, they are desirable difficulties.

Students can make use of spaced practice in their everyday studying. For example, when using flashcards, a student could study one big set of flashcards spaced across days. Yet using small sets of flashcards and studying them all at once is a common practice, and my subjects thought it worked. Nonetheless, they learned far more GRE words when the flashcards were spaced out (Kornell, 2009b).

What about more complex topic matter, such as learning a new category by observing examples? When learning a painter’s style, or what a certain butterfly species looks like, studying the same category repeatedly seems like a good idea, because it makes it easier to notice similarities. That’s what our participants thought, at least, but in fact interleaving different artists or species is, again, far better (Kornell & Bjork, 2008a). The same is true for older adults (Kornell, Castel, Eich, & Bjork, 2010) and three-year-old children (Vlach, Sandhofer, & Kornell, 2008).

Retrieval is the ultimate goal of learning, but it’s also a potent way to learn in its own right. It makes sense that retrieval practice enhances later retrieval, but people often see re-studying as more effective than retrieval practice (Kornell & Son, 2009). It’s confusing because we often see mistakes as a sign that we aren’t learning, and if you don’t retrieve you don’t have to make mistakes. In fact, really hard tests really help learning. In fact, tests can enhance learning even if one cannot retrieve the answers on the test (Kornell, Hays, and Bjork, 2009; Richland, Kornell, & Kao, 2009).

How people think learning works
The study decisions we make have practical consequences, although few of us ever receive training in how to study (Kornell & Bjork, 2007). One common real-world decision is whether to “drop” (i.e., put aside and stop studying) flashcards. When people decide to drop flashcards, it can actually make them learn less because they become overconfident (Kornell & Bjork, 2008b). People also tend to misunderstand the value of spacing and retrieval practice. Not all study decisions are bad, though; for example, people tend to study the information closest to being learned (Metcalfe & Kornell, 2005).

Judgments about memory (e.g., I knew that!) and beliefs about memory (I can’t remember names) are surprisingly disconnected from each other. They’re also disconnected from the actual memories themselves. For example, the easier information is to process, the better we think we know it. As a result, we think increasing a word’s font-size makes it more memorable, which it doesn’t (Kornell, Rhodes, Castel, & Tauber, 2011). On the other hand, we don’t appreciate the impact of what’s going to happen in the future, because the future doesn’t impact ease of processing right now. This leads to a stability bias: people fail to predict that they’ll learn from future studying (Kornell & Bjork, 2009) and they underestimate future forgetting (Kornell, 2011).

Animal cognition
Non-human animals benefit from desirable difficulties; for example, retrieval enhance learning in monkeys (Kornell & Terrace, 2007). Monkeys also make accurate metacognitive judgments; they will bet more when they’re sure of a memory than when they’re unsure, and they’ll also request additional information when uncertain (Kornell, Son, & Terrace, 2007).