Matthew Hays, Ph.D., VP Research and Analytics at Amplifire

Insights from my recent presentation at Psychonomics on why learners can’t trust their instincts about what works.

People are really bad at managing their own learning. A big reason for that: They mistake how easily something comes to mind right now for how easily it’ll come to mind later.

Imagine two people getting ready to give their own conference presentations. Suppose Beth practices her talk once a day for each of the four days before the conference, and suppose Bill practices his talk four times in a row on the day before the conference. At the end of his marathon practice session, Bill’s going to be feeling amazing. But the next day, Beth’s talk is going to go more smoothly. Bill’s going to be scratching his head, wondering how practice went so well, but the talk went so poorly.

Two things are at play here. One is the spacing effect, which is one of the most powerful cognitive phenomena on record. It’s the finding that spreading your practice sessions over time massively increases how much you learn. (Side note: If you don’t do anything else to improve your memory, use spacing!) 

But the other thing going on here is that Bill is mistakenly interpreting his great performance during practice as evidence of great learning. What really happened: Beth set her practice sessions up to be more difficult while she was practicing, but to produce better learning as a result. Bill set his practice sessions up so that he would improve as rapidly as possible…but at the cost of tomorrow’s talk.

Phenomena like the spacing effect are called desirable difficulties, which are challenges during learning that create better long-term retention. But they’re unintuitive; people don’t know about them and instinctively avoid them, like Bill did. Instead, they gravitate toward what Dr. Dillon Murphy and I call evil easies: conditions of learning that make you think you’re learning a ton but then leave you hanging.

There are all sorts of ways that people’s misconceptions about their learning cause them to mismanage it. Dr. Murphy and I looked at a particular decision people make, both during their formal education and in the corporate world: Speeding up the playback of instructional videos. In some previous research, Dr. Murphy found that only about 15% of undergraduate students watch pre-recorded lecture videos at normal speed; the rest accelerate videos to various extents.

Their decision to speed up videos becomes interesting in light of some other data that Dr. Murphy collected. He found that speeding up videos, all the way to double speed, did not impair learning! But when he asked his research participants about it, he found that 42% of them thought that normal speed (1x) was best for learning. …But wait, we also know that only 15% of students watch at 1x speed. Clearly, some students are speeding up their lecture videos even though they think it’s bad for their memory.

Here’s the twist: In addition to selecting the “worse” option for their learning, they were also wrong about the trade-off! Because speeds up to 2x don’t hurt learning, these students were making the right practical decision saving time without sacrificing comprehension – while being completely mistaken about what they were doing to their learning. Two wrongs made a right!

What This Means

At Psychonomics this year, Dr. Murphy and I presented these data as yet more evidence that learners are lousy managers of their own learning. They’re often wrong about how learning works, and they also often fail to act in accordance with their own beliefs. In other words, they have fundamentally flawed metacognition.

Our presentation was remarkable not because we demonstrated learners’ metacognitive failures – plenty of researchers have done that – but because in this case two metacognitive errors canceled each other out. Most times, learners aren’t so lucky.This is where Amplifire comes in. We can’t rely on learners’ intuition to guide their study strategies…but we can design systems that guide learners toward empirically validated, efficient learning behaviors. Well-designed tools like Amplifire can turn accidental success into consistent success.