Guessing before you know the answer creates suspense: a cognitive receptacle for the answer + emotional engagement. When the answer comes along, you grab it, like an enzyme grabs one of its raw materials, because it fits. And it's a relief.
When studying at uni, I took a past exam paper under exam conditions, before doing any study: it wonderfully concentrated the mind on what I needed to learn. (I got the top mark, with a big gap to the next person).
When teaching at uni, I presented a puzzle at the start of the class, that didn't make sense until the material in that lecture had been heard. The entire class came up to me at the end to check what the answer was.
Learning through trial and error is so damn useful because that's how our brain is designed to learn.
Whenever we make a prediction, dopamine neurons in our brain start to fire, triggering a feeling of expectation. If we predict right, we get a surge of dopamine and some positive reinforcement. If we get it wrong, the dopamine neurons stop firing which, aside from making us feel like crap, serves as a signal that the prediction needs to be revised. Normal learning operates by consciously acquiring information and hoping it eventually filters down into the subconscious. Trial and error, on the other hand, puts a direct line into our subconscious - it's much more efficient.
Sometimes I wish educators studied some Machine Learning and similar things.
Not to detract from the interesting finding, but it sounds to me like someone discovered that in supervised learning it is advisable to use a negative corpus as well :)
Not only that, but making mistakes is often the best way to learn how to do something correctly.
People who don't make very many mistakes are usuall also people who don't try to do anything new, so they're not giving themselves a chance to learn to begin with. Learning to make and accept mistakes is one of the best lessons anyone can learn; it helps you to keep your mind open so that you CAN learn.
Learning by mistakes is a good way to learn im sure of it.
But this article on page 1 covers more about learning where you have a teacher who already knows the material, it would be far harder to use this failing technique if you are creating the material to begin with.
Minus the above I can relate to learning by failing.
For those who use the command line in unix. I hardly can ever remember the command if I grabbed it off google. When I attempt, fail, look it up knowing what I am after I always find myself more often than not still with the correct knowledge.
I do feel this knowledge is sort of common knowledge or more like "yes, that is what i would have said if i ever put it in words". Needless to say it is good to read these types of obvious thoughts as it just aids the life learning process.
"In fact, they found, learning becomes better if conditions are arranged so that students make errors."
Well, then, there's no better learning environment than computer programming!
"People remember things better, longer, if they are given ... tests at which they are bound to fail."
No pain, no gain.
I used to wonder why a damaged body part is painful for so long after it has 'made it's point'. After I'd lived long enough to forget several painful lessons, I understood: the longer the pain lasts, the longer the memory lasts.
Makes me think scars are evolution's way of helping us to remember.
The approach they outline is nothing new. It's how I developed mathematical proficiency 20 years ago, following the advice in Mistakes and how to find them before the teacher does.
When studying at uni, I took a past exam paper under exam conditions, before doing any study: it wonderfully concentrated the mind on what I needed to learn. (I got the top mark, with a big gap to the next person).
When teaching at uni, I presented a puzzle at the start of the class, that didn't make sense until the material in that lecture had been heard. The entire class came up to me at the end to check what the answer was.