A group of scientists set out to study quick learners. Then saw they don't exist

"The fastest quarter of students improved their accuracy on each concept (or knowledge component) by about 2.6 percentage points after each practice attempt, while the slowest quarter of students improved by about 1.7 percentage points."

That's a small difference in a single incremental component, but it represents a difference in learning rate of a whopping 35%; over ten practice attempt, the fast learner would've accumulated 26 percentage points of improvement, the slow learner only 17 percentage points. And there is a bit of a compounding effect, as the accumulated advantage of the fast learner would unlock progress on additional components that the slower learner lacked enough background for.

Add these two effects over the thousands of components that would make up a curriculum throughout a K-12 career and it wouldn't surprise me if you had top learners accumulating over twice the progress of their slower learning peers, and much of that knowledge would be in higher value components inaccessible to the slower learning students (eg, learning calculus and AP biology over geometry and basic science).

I was regularly hamstrung by curriculae because I learned quickly. Schools chose to throw a bunch of busywork at me instead of deepening or questioning my understanding. "Your peers aren't on that yet." I don't care what my peers are doing. They made their decision about me years ago. If school can't be intellectually challenging then what's the damn point.

It sounds like students were pre-tested, given instruction, and then post-tested. The researchers assume that students who did better on a pre-test were solely leveraging prior knowledge. The researchers do not consider that higher pre-test scores could have been due to the ability of some students to discern the correct answer from context, analogies, or first principles. Students who are able to do so would be seen as "quick learners" by their teachers and peers — though not by these researchers. These students would not need as much time to master all of the content in a lesson, since they would grasp perhaps 50% more of it than their peers right out of the gate.

The researchers focus on the fact that these students appear to learn the additional bits at roughly the same rate as their peers. But of course, if they have much less left to cover in a given lesson, then they will be done with the lesson well before their peers. By most definitions, this makes them 'quicker learners', assuming their initial performance was not based solely on prior knowledge, but was also based on their ability to figure out some of the material from the initial questions themselves.

Also, one would assume that the last parts of a lesson are the hardest to learn, so a student who understood 90% of the lesson at first might still be learning the last 10% faster than a student who came into the lesson understanding 50% of the lesson (when such student gets to the 90% point). However, the student who student who starts at 90% might appear to learn at the same rate as the other student because the other student is getting the 'low-hanging fruit' and moving from 50% to 60%. This should not be considered equivalent to moving from 90% to 100%.

It's very frustrating with these studies because they don't attempt to normalize anything at all. Imagine teaching derivatives and finding that some students learn faster than others. Their excuse? Oh it must be because they did fractions when making pancakes.

One of the easiest ways to create a replicable study, which they won't ever do, would be to place students who have no experience typing (which is common for every generation) and see which ones learn fastest. If you are bothered by this then we can do learning piano, filtering out any student which may have musical practice ahead of time. It would be easy to find. Hypothesis would be that those with higher IQ have a higher rate of mastery.

Extraordinary claims require extraordinary evidence, and I'm not sure that single study meets the bar.

It is certainly a valid issue to measure starting point as well as ending point to measure learning rates, but to claim that this entirely accounts for apparent differences in learning rates such that there are no differences exceeding 1% just strains credulity.

Just looking at my own learning rates on different fields I can see myself picking up some things far faster than other things, and the reverse in friends and family (e.g., quickly picking up things I'm slower to learn).

Seems more like an intriguing start to more research rather than the conclusive tone presented in the article.

Reality is that I had been programming in C++, fortran and DSL for over a decade.

Similar examples from colleagues in college and schoolmates. People with strong foundations seem to learn faster, but the reality is that there is past effort that you don’t see.

Unfortunately now I can't find where I read that despite extensive searching. I want to say it was attributed to Maslow.

Anyway, this article claims the exact opposite by saying people learn at the same rates.

Actress Marilu Henner, best known for her role as Elaine Nardo in the hit sitcom Taxi, has a highly superior autobiographical memory (HSAM), a rare condition shared by only 100 people worldwide.

I suspect those 100 people would be able to learn certain material faster than others.

I'm also curious about people who have learned mnemonic tricks. I've seen people memorize the order of the US Presidents in just a few minutes once they are taught a memory trick...

But yeah, the point of the article, in my mind, is that nothing beats investing a lot of time in especially in early childhood exposing people to all kinds of learning opportunities.

that explains the smell.

I wouldn't consider myself an expert in non-natural science, even though I do understand some of it to an extent.

So I could take this at face value, it's quite plausible that nobody has ever been fast.

This fully explains the confirmation of different learning rates universally observed over the millennia, we're all just different degrees of slow.

The underlying study is not about just rate of learning, but rate of learning under favorable conditions.

The article describes where the data comes from:

> In particular, we model learning using 27 datasets with over 1.3 million student performance observations from 6,946 learners in 12 different courses ranging across math, science, and language learning, across educational levels from late elementary to college, and across educational technologies including intelligent tutoring systems, educational games, and online courses

And the authors argue this is favorable learning conditions because of providing things like immediate feedback on errors, etc..

Lots of room for nuance, but "favorable learning conditions" is key here.

https://www.pnas.org/doi/10.1073/pnas.2221311120