You’re the understudy for the lead in the school play, and the lead just got sick. It’s the midnight before the show. What to do?

90s sitcoms had an answer: listen to the lines while you’re sleeping. When you wake up, you’ll know the lines perfectly and be brilliant up on stage. Of course, in the sitcom this sleep learning method never turns out quite right, and high jinks soon ensue.

In real life, the method wouldn’t work at all. Sleeping well, of course, is quite vital for learning. Listening to something while you sleep and expecting to learn it: not so much. But it’s no more ridiculous than a lot of crazy things that we believe about learning.

Lectures, Baby Einstein, and “Whole-Brain Learning”

The widespread belief in the Baby Einstein products, for example, which initially pitched listening to classical music as the way to create genius kids, seems laughable. How, exactly, was listening to Mozart or classical poetry supposed to make you smarter? This was just the education world’s version of the “profit!” meme:

Step 1: Listen to Mozart.
Step 2: ?
Step 3: Be a genius!

But consumers (and even some educators) accepted this chain of reasoning as totally plausible. Classical music is associated with smart people. Classical music seems complex. So let’s have kids listen to classical music. But this indirect approach just doesn’t work. If kids listen to classical music, they’ll probably be able to distinguish pieces of classical music a little bit better than kids who haven’t. But that’s about it. They’re not going to get better at math. They’re not going to get better at reading. Because they’re not doing those things.

The “whole-brain” method of teaching that has gained recent popularity has students clapping, gesturing, and endlessly repeating catchphrases while they’re trying to learn. Proponents claim that doing these things “engages the whole brain” making learning more effective. This is just a variation of the “we only use 10% of our brain” meme. Brains don’t work like that. Our whole brain is in use all of the time. Distracting part of it by doing something else is not going to help. Effective ways of using our bodies to facilitate learning exist (“embodied cognition” — for instance, learning proportion through feedback on hand movements), but these approaches focus student attention on what they’re learning — they don’t ask students to multitask.

Of course, the same failure of imagination comes into play with thinking that lecturing students is an effective way of teaching them. Sometimes teachers ask, “How can I make my lectures more effective?” But this is the wrong question. Lectures provide few learning opportunities. They don’t give students practice at developing their skills, and provide no feedback to students about whether — and what — they’ve learned. There are ways to improve lectures: use simple, direct diagrams to illustrate your ideas; deliver material at a pace that students can handle; avoid distracting or irrelevant details. But the best way to improve a lecture is to make it stop being a lecture: get students involved more actively in solving problems and thinking through key ideas.

Does it work?

Often, conversations about new learning approaches center around whether the approach has been proven to work. Have studies on its effectiveness come out? The companies selling “innovative” learning methods will invariably say that their approaches work: “research” demonstrates that the approach really improves learning (often the company’s own research that hasn’t yet been subject to wider scrutiny).

This has been the approach that “brain-training” games have often taken. Lumosity developed games to improve general cognitive skills like working memory, attention, executive function, and so on. Lumosity’s own research supported their claims of mental improvement. But third-party research on their games showed otherwise. When Lumosity games were tested against a game simply designed to be fun — Portal — the game designed to be fun improved things like cognitive spatial skills. Lumosity games did not improve any aspects of cognitive functioning. The company ultimately paid $2 million to settle claims of false advertising by the FTC.

As with fake news, people can generate more crazy ideas about learning than can be empirically tested. And, because testing takes time, unbiased research on whether the approach works in practice invariably comes out after a learning approach has already established itself in the marketplace, instead of before. In the absence of research evidence, we evaluate novel learning approaches in other ways. Do students like it? Do teachers like it? How many school districts are using it? How much capital has the company raised? Are my friends using it? Is there buzz? None of these are bad things, but they aren’t good measures of whether an approach actually works.

I suggest a different way of evaluating new learning approaches — especially those that you have to pay money to use. Is it plausible that things could work that way? Forget the brain-related words that companies use. Does clapping a lot seem like a way to help students improve how they multiply fractions? As with Baby Einstein, there’s just no known mechanism that links the activity (clapping) with the learning goal (multiplying fractions). The mechanism that whole-brain supporters propose — that the approach uses “all” of our brain instead of traditional approaches which only use “some” of our brain — comes straight out of science fiction.

The plausibility test

So, how to evaluate whether a learning approach can plausibly help students learn? Let’s start with four principles, drawn from many decades of learning research:

Learning takes effort. Do not trust learning approaches that make claims about learning effortlessly. Good learning approaches structure effort; they don’t eliminate it.

Learning one thing doesn’t make you better at something else. Culturally, we associate certain skills with intelligence. Good at chess? You’re probably “smart” (and therefore good at lots of other things). But being good at chess is just about being good at chess. Expert chess players don’t have better memories or attention spans. They have better memories for chess; they have better attention spans for chess. Because they have developed knowledge structures that help them to analyze chess positions. If students aren’t actually practicing the skill that you want them to learn, they’re not going to learn it.

Distraction is bad. If you’re purposefully distracting students while they’re learning, they will not learn effectively. Students’ attention needs to be focused on what they’re trying to learn.

Feedback about how to improve is vital. If students aren’t getting feedback about how well they’re doing — and how to do better — they’re simply not going to improve.

Learning is an immensely complex topic. But when learning approaches violate one (or more) of these principles, we should all be skeptical.

If you practice playing Lumosity games, you will get better at playing them. You might even feel good about yourself. But instead of learning public speaking or calculus or painting you’ve learned how to count a bunch of cartoon birds.