Metafact Review: Learning

Learning is fundamental to human progress. Babies for example, are constantly absorbing the sights and sounds around them - taking queues from their environment. Children begin to communicate without their parents teaching them - they just seem to spontaneously start. Teenagers then engage in formal learning programs at schools and universities. As adults, we continue to learn. Whether it's at work or home, from learning how to cook Mexican food or another language. Yet life is short and time is precious, so everyone is trying to figure out how to maximize their learning progress.

It turns out that there are many, many types of learning. There's verbal learning, where we learn words, symbols or pictures associated with language. There's motor learning, where we learn how to ride a bike or drive a car. And in conceptual learning, we learn about concepts (you guessed this). When you learn that the Moon is closer than Jupiter, or that the capital of Burkina Faso is Ouagadougou, it is this kind of learning that you are doing. And so on.

With such a diversity of ways in which our brains can acquire new skills, behaviors or knowledge, it is no wonder that there is a fierce debate among experts on how to help people learn better. This is especially the case in educational settings, where different learning theories and proposals are constantly put forward, many get quickly debunked.

The good news is, this month we asked scientists to share the facts about learning. What is the difference between all types of learning? Can you learn new things when you are old? Should schools move to teaching outdoors? Can you learn a new language just by using an app?

Here's what we found...

Read the full review


40 million: Number of active users on language app Duolingo.

2%: Percentage of the UK population with a learning disability.

1 in 10: Estimate of people with dyslexia in the UK.

13: Average years of formal education in Australia, US, Canada or Japan.

1.5: Average years of formal education in Burkina Faso (the lowest anywhere).

52%: Percentage of UK families where children struggled with homeschooling during Covid.

14: Minutes of study per day by an average person in Finland.

Background: Types of learning


On the back of many London buses, there is a banner that says “watch your speed, your child does”. The goal is to make you be more mindful of your speed. For psychologists, there are three main ways in which we can learn new behaviours. The London campaign refers to the most basic one, whereby we observe others and imitate what they do. That’s observational learning.

Classical Conditional

In the second type, we link two unrelated events, like when hearing people in the kitchen makes you feel hungry. This example may ring a bell (pun absolutely intended), as it is similar to Ivan Pavlov's famous experiment, where dogs learned to salivate in response to a bell tone. Because the bell rang every time the dogs were presented with food, the animals learned to associate both events. Of course, there is no obvious causal link between a bell being rung and food becoming available. However, in this specific context, the animals become conditioned to link both events. This form of learning is called classical conditioning and it’s very much present when you hear a baby cry, in traumas or even during sex. When you once felt ill after eating something unusual and now feel nauseous if you eat it again, that is classical conditioning in action.

Operational Conditioning

There is yet another form called operational conditioning. This occurs due to the consequences of actions. In other words, when a response is reinforced upon rewarding the subjects, or weakened if we punish them. This is a standard principle behind behavioural training and also underlies education systems to some extent. Here, unlike in classical conditioning, the response is in our hand and reinforcement or punishment help (condition) us to learn.

American psychologist B.F. Skinner, now considered the father of operational conditioning, ran another classical but somewhat less famous experiment to prove it. He did it on pigeons and rats, getting identical results. The animal is put in a cage (Skinner’s box) where one of the walls has a lever. If it accidentally bumps on the lever while exploring randomly, a food pellet appears. The response is thus reinforced, and the animal learns to directly press the lever when it wants food. It also works for avoiding unpleasant stimuli too.

In a separate experiment, the cage floor has an electric grid and an electric current is applied just after a warning light goes on in the cage. By chance, the animal might trip and press the lever, which switches off the electric current. Over time, it will learn to race to the lever as soon as the warning light goes on to prevent the electric shock. Skinner’s principles are now well known by marketing gurus, used in slot machines and social media algorithms to exploit this form of learning.

Disclaimer: these are all happening in the same brain, at the same time. It is true that different brain areas can be involved in different tasks (and thus learning of tasks). For instance, the hippocampus is crucial for spatial learning, and the cerebellum on the back of your brain controls motor coordination so it plays a key role when you learn how to drive. However, forms of learning intertwine, so some degree of discrepancy remains on which underlies what sort of tasks - and it is probably a combination of them that does the job.


Do we need tests to learn?

67% Negative via 9 experts

Tests are a form of operational learning. If the entered answer is correct, we get rewarded with points. If it is not, we lose that reward. Most education systems are based on the premise that tests are a valuable tool not just to assess but also to boost learning. One might wonder, however, whether children (or adults) really need them.

“Tests and exams are one way to assess pupils’ learning,” writes Prof Dominic Wyse from the Institute of Education of University College London. He stresses, however, that they are not our sole option as “there are other ways of assessing pupils’ learning that give formative information to teachers.” He argues that “optimal learning happens when teachers assess children’s learning then base their teaching on their assessments of children learning.”

Although well-designed tests (we will come back to this later) do help people learn faster, our experts agree that tests are not required for learning, strictly speaking. According to Prof Wyse, “we all learn some things without any assessment at all so the literal answer to the question is no we don’t need tests or examples to learn.”

Having said that, there may be differences between children and adults. “As an adult with good self-regulation and metacognitive skills, I learn all the time without tests,” argues Prof Diane Grayson from the University of Witwatersrand in South Africa. However, for the younger ones, learning can be improved by “structured opportunities to demonstrate their learning and to get useful feedback, such as can be provided in a well-designed and carefully graded test,” she adds.

But not all tests are useful. Tests can be a bad idea, says Prof Grayson, if they do not support the learning goals or when constructive feedback is not provided for the learner. In these cases, tests “may, in fact, undermine deep and meaningful learning if the test questions are too divergent from the learning outcomes or the feedback is cursory or, worse still, demoralising,” she argues.

So what is a good test? According to Dr Ayşenur Sağdıç from Georgetown University, it should do four things:

“1) provide insights and feedback on what has been learned and what has yet to be learned.

  1. motivate learners to continue improving their knowledge or skill set.

  2. motivate educators to implement optimal instructional practices, also known as positive washback.

  3. allow education stakeholders to make actionable and systematic recommendations in terms of program needs and goals.”


Can you really learn a language with an app?

40% Affirmative via 5 experts

A lot of us have at least attempted to learn a foreign language using an app. They provide an easy, fun and mostly free alternative to traditional learning methods. For education experts, these apps provide an opportunity to improve teaching using an innovative tech-based approach. However, some are skeptical that they can help us learn a language beyond the basics.

“Apps can be an effective and convenient way to get started with a language or maintain your current levels of knowledge,” writes Prof Dan Isbell from the University of Hawaii. “Apps are fairly good at teaching vocabulary and basic grammar structures, which are foundational to language skills,” he adds. “There is some evidence that students using only apps like Rosetta Stone and Duolingo perform at least as well on language tests as students in classroom settings,” writes Prof Edith Kaan from the University of Florida. “Especially learners using gamified apps like Duolingo say they enjoy learning more and are more motivated,” she adds.

Several experts pointed out that learning a language does not mean one thing, however. Instead, it forms an array of diverse tasks, each underlied by different types of learning. In this line, we should have reasonable expectations when learning a new language using only apps. “Most apps are limited in how far they can take you, and at some point, you'll need to go beyond apps to hone communicative skills and develop broader, richer knowledge of a language,” says Prof Isbell.

Learning associations between a list of words in your mother tongue and their translations in a foreign is one thing, but there is much more to language learning than that - including its social dimension. Apps are not great for learning certain aspects of a language. “Especially conversational skills and knowledge about the culture and history of the country where the language is spoken,” argues Prof Kaan. Dr Sağdıç agrees: “Language competence involves not just knowing the explicit and discrete rules of the target language but also the implicit rules, the pragmatics (language use in its social context),” she says.

Either way, not all platforms are the same and various factors can affect our gains. According to Dr Sağdıç, “one of those factors is the app itself” as different apps include different features (such as VR or real native speakers on the other end), she explains. Other factors include “the length and the frequency at which users engage with the app, motivation to learn the language, users’ age,” she says. Your environment also plays a role, she says, as you are more likely to engage in the learning if you receive offline classes or if you live in an area where the language is widely spoken.


Can you learn in your sleep?

57% Affirmative via 7 experts

Picture this - going to sleep while an audiobook is playing and waking up with all that knowledge in your head. Wouldn’t it be wonderful? Associative learning may be behind this dream. According to a 2018 review on the topic, when we associate a sensory cue (for example, an odor) with a piece of information presented alongside it, memory of that information can be strengthened during sleep if we are presented the linked sensory cue. This is a new technique, called Targeted Memory Reactivation (TMR). “This concurrent presentation supposedly creates associations between the task being learned and the unrelated stimulations,” explains Prof Itamar Lerner from the University of Texas. “Evidence from animal research suggests that these repeated presentations reactivate memories related to the learned task, leading to their strengthening,” he adds. While this is an interesting experimental paradigm, unfortunately, it clearly has little to do with real-life scenarios. On current evidence, being able to study for exams just by playing lecture recordings during your sleep remains a fantasy.

For Prof Lerner, it would be naive to believe that we can learn complex information as we sleep. Yet, “simple strengthening and reorganization of previously learned materials may be possible”, he writes.

In line with this, Prof Michael Thomas from the University of London Birkbeck explains that sleep is fundamental for learning. First, it helps clean up the brain (from chemicals that build up during the day). Second, it helps consolidate the learned information. “Consolidation involves strengthening the brain connections of knowledge or skills learned in the day, and also transferring information from a structure called the hippocampus, which stores episodic or autobiographical 'snapshot' memories during the day, to the cortex, which stores conceptual knowledge”, he writes. This process needs the brain to be offline, though, “so you couldn't view the brain as an effective 'classroom learner' while it's asleep” according to Prof Thomas.


Does the ability to learn decrease with age?

63% Affirmative via 8 experts

Brains, like all our other organs, age. Our ability to form, store and retrieve memories changes as we get older, and this is particularly dramatic in people with dementia, where serious deficits can exist. Children are commonly said to learn really easily, but where does that leave the rest of us? Are we able to learn new things when we get old?

“A vast body of research shows clearly and crisply that the ability to deal with novelty and solve new problems peaks in the late teens and early 20s before beginning to descend rapidly in middle age, and continues to fall dramatically with increasing age,” writes Prof Alan S Kaufman from Yale University.

However, a person’s ability to store information does not shrink until much later. “The ability to accumulate knowledge, whether in school or by other cultural experiences, increases with age until a person is in their 60s (when level of education is controlled), and doesn’t start declining until age 75,” he adds.

In other words, someone in their 70s will probably struggle to learn a new skill, such as playing a musical instrument, compared to someone in their 20s. But if both of them had learnt how to play it in the past, we might be in for a tie. So it all depends on whether we are already familiar with the information to be learned, according to Prof Kaufman. “If it is something novel, then the ability to learn the new material, or the new skill, will decrease dramatically with increasing age. If it is familiar then there will be no decrease in learning ability with age, and there may well be an increase,” he argues.

Prof Marvin Formosa from the University of Malta notes that older learners might even have an edge. “Indeed, older persons actually possess a number of compensatory factors - such as the integrity of crystallised intelligence, the accumulation of knowledge and experience, the persistence of curiosity, and the ability to put new information in a highly meaningful context,” he explains. Prof Gavin Brown from the University of Auckland agrees with this. “They have abilities into which new knowledge can be structured—it’s why older and wiser people with knowledge of the past add value to decision making,” he notes.

However, Prof Formosa cautions that it is hard to say whether any differences in learning ability with age are really due to brain ageing, as many other factors could play a role. “Increasing birthdays may be the least contributing variable to such developments next to social crises, personal tribulations, health complications and environmental changes”, he explains. Studies comparing groups of younger and older people are also of limited power, he adds, given the huge variability between people and between experimental settings across studies.

Quick Answers

Do you remember things better if you say them aloud? Speaking to yourself makes it feel personal - which helps, according to research - and adds a more proactive component to learning.

Do we just use 10% of our brain? No. Although only a fraction of neurons are active at any one time, we use all of them to process information and learn.

Does learning lead to physical changes in the brain? Yes. Learning how to juggle, for example, increases the amount of gray matter in a brain area involved in visual learning. This change was transient, however, so better keep practising.

Are ‘forest schools’ better for teaching? Still uncertain. Teaching outdoors can help children engage and pay more attention, and there is some evidence it can improve performance. Yet, it is not the most important factor for learning and the trick seems to be making teaching more student-centered.

Is single-sex education better? Quite a bit of research has been done on this. Despite evidence that academic performance can improve in single-sex schools, others claim that an analysis of the wider picture reveals no academic advantage - and indeed worsening of sexism and gender stereotypes.

Is there a correlation between grades and intelligence? Yes, smarter people tend to do better at tests. However, many other factors are involved including socioeconomic background and learning disabilities.


Does a learning disability strongly impact life quality?

The social provision for those with learning disabilities varies widely across the world, and with it, the health and quality of life of those with learning disabilities (or intellectual disabilities as it is known internationally). A persons quality of life is a multidimensional concept but often includes health at the core. Whilst learning disabilities is an umbrella term including many different developmental conditions and genetic disorders, this heterogenous group do share various health experiences/risks.

People with learning disabilities often experience a burden of long-term health needs including high rates of epilepsy, mental ill-health, gastrointestinal disorders, diabetes, obesity, physical disability and polypharmacy (taking 5+ medications). People with learning disabilities die younger than people without learning disabilities, on average, up to 20 years earlier, often from reasons considered avoidable.

There are undoubtedly significant inequalities and injustices faced by those with learning disabilities, but these can be changed, improved and prevented. More equitable and person-centred healthcare can reduce the inequalities faced by people with learning disabilities, enabling each individual to have good health and wellbeing. As with every person there are gradients and many factors that influence our wellbeing; people with learning disabilities are no different, like other minorities/marginal groups it is societies and systematic influences that shape inequalities and inequities.

Dr Laura Ward from the University of Glasgow


  • There are many different types of learning along with the brain structures and processes that power them.

  • Language learning apps are a useful tool to improve, but you would be better off to practice outside the app too. Find yourself a conversation partner, perhaps?

  • People differ greatly in how (and how much) they learn, too.

  • Old dogs can learn new tricks. It just depends on the type of learning we’re looking at, but old people can still learn - particularly things they are familiar with.

  • There is no point in playing your podcast while you sleep. Just enjoy it over a cuppa before you turn the lights off!

  • No clear evidence that teaching outdoors or single-sex schools produce better academic or learning outcomes