Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Effective learning: Rules of formulating knowledge (1999)

supermemo.com

Post reply on HN
Print view View original

21–30 of 37 posts

Re: Effective learning: Rules of formulating knowledge (1999)

#21

Post by theoh »

guizzmo wrote: post #6

The first rule "do not learn what you do not understand" is nonesense in the context of learning mathematics/physics/hard sciences. If you understand a subject, then it means that you have learned it effectively. In my experience with teaching at university, I have found that the wish of students to understand before learning is actually a great barrier to comprehend a subject. In order to build intuition on a subjec…

In the hard sciences, learning by rote, or just learning to manipulate symbols according to rules, is not a good long-term strategy.

http://www.physics.emory.edu/faculty/weeks//journal/elby-ajp...

Re: Effective learning: Rules of formulating knowledge (1999)

#22

Post by rdslw »

mistermann wrote: post #13
rdslw wrote: post #9

If you're wondering if it's (supermemo) worth your time, let me share with you my 10 (ten) years supermeme learning conclusions. I got supermemo advanced english course (20k phrases) around 2004 as a present and started it in 2005. Initially I thought it will take 2..4 years. Exactly 10 years later, on September 2015 I was at 19k. I reached last 20k phrase (but not finished the course) around spring 2016. I ended doi…

Now that you're done, why do you think your initial estimate was so far off?

Two things: a. underestimating 'compound interest' (if I may say so) of repetitions from past material you must do in the future. As more and more material is done, there are more repetitions you must do every day, simply because you're forgetting some of the material. So simple '20k / X new words per day' gives BIG error. b. "I had multiple one day slips" is another one, especially as you quickly "learn" that the more new ones you do, bigger daily repetitions number follows :) So quite often I was just doing repetitions to reduce burden on me.

Guessing from my memory, average repetition number over that 10y would be around 100 per day.

BTW: You can start with smaller db. e.g. I started in the middle doing Russian one and used a 5k one.

Re: Effective learning: Rules of formulating knowledge (1999)

#23

Post by rdslw »

versteegen wrote: post #16
mbrookes wrote: post #14

Earlier quoted context omitted.

This is not a dig at the parent - they clearly have commitment and drive, but since they're promoting SuperMemo rather than commenting on the post, let me just say that this: "I would did it again!!" or this: "Of course better english command came also with the course" (among a choice of several examples) should tell you everything you need to know about the effectiveness of SuperMemo after 10 years of learning.

But 12.5 minutes a day for 10 years is only 761 hours. The purpose of SuperMemo is to optimise for total time spent. (Of course, the specific course/material might not have been the most effective.) But I guess that argument doesn't hold water, since rdslw is surely also exposed to a lot of other material in English.

For sure I was.

And yes, this specific course rather wasn't the most effective one (but probably best at that time). There were multiple problems with this supermemo db: many simple phrases I already knew, but also lot of words you never use (afterwards). Also the biggest problem were circular errors (as I called them): A>B but also A>C or B>D, where my brain was struggling with one knowledge fact replacing the other one and again..

Re: Effective learning: Rules of formulating knowledge (1999)

#24

Post by emilga »

cle wrote: post #5
nomurrcy wrote: post #4

I've been using Anki for the past year while learning German and it has been (seemingly) very helpful. I doubt I would have acquired the vocabulary I have in the past year without it. I'm not sure if this is simply due the fact that it helps make studying a habit or that SRS is as beneficial as claimed. I have found it useful enough that I'm extending its use to other subjects I'm interested in. Has anybody used Anki…

SuperMemo has a lot more advanced features (e.g. incremental reading). But the tradeoffs are pretty significant, in my opinion. Anki's decisive features for me are its mobile apps, its simplicity, and its open source license. I've been using Anki for 5 years now and have about 10,000 notes. I don't use it for foreign language learning. I would not feel comfortable investing that much effort into a closed-source platf…

How are you able to keep 10000 notes in Anki organized? With that many notes, how can you tell whether something you want to remember is already in your notes or not?

It seems to me that the Anki card browser is quite bad at this, but maybe I'm just bad at using it!

Re: Effective learning: Rules of formulating knowledge (1999)

#25

Post by rdslw »

mbrookes wrote: post #14
rdslw wrote: post #9

If you're wondering if it's (supermemo) worth your time, let me share with you my 10 (ten) years supermeme learning conclusions. I got supermemo advanced english course (20k phrases) around 2004 as a present and started it in 2005. Initially I thought it will take 2..4 years. Exactly 10 years later, on September 2015 I was at 19k. I reached last 20k phrase (but not finished the course) around spring 2016. I ended doi…

This is not a dig at the parent - they clearly have commitment and drive, but since they're promoting SuperMemo rather than commenting on the post, let me just say that this: "I would did it again!!" or this: "Of course better english command came also with the course" (among a choice of several examples) should tell you everything you need to know about the effectiveness of SuperMemo after 10 years of learning.

Nah, rather about me (sloppiness + using English only from time to time) and (this is big for grammar/style) that my native language is quite far from English in terms of lexical distance, and this creates a lot of problems when your brain (thinking in native language) translates word-by-word to other language.

Especially that I said 'better' not best/perfect etc. As someone commented, SR is not really good for grammar.

But thanks for spotting my mistakes :)

Anti-disclaimer: I'm not in any way connected to supermemo, having profits etc. It's funny that my post could have been seen as a promotion :)

Re: Effective learning: Rules of formulating knowledge (1999)

#26

Post by drieddust »

guizzmo wrote: post #6

The first rule "do not learn what you do not understand" is nonesense in the context of learning mathematics/physics/hard sciences. If you understand a subject, then it means that you have learned it effectively. In my experience with teaching at university, I have found that the wish of students to understand before learning is actually a great barrier to comprehend a subject. In order to build intuition on a subjec…

> So my advice: don't try to understand. Do, and do again untill you have learned. Not the other way around.

This is so so true so some level I think our desire to understand everything beforehand is just a lazy strategy. Probably either to learn only what is easy or what is useful. Sense of wonder and discoverablity which makes putting the effort in a joyous activity is gone from us. Our profit mazi

On the contrary Children just keep practising until they learn. I wish I can go back to stage where my 2 years old daughter is.

Re: Effective learning: Rules of formulating knowledge (1999)

#27

Post by Eerie »

emilga wrote: post #24
cle wrote: post #5

Earlier quoted context omitted.

SuperMemo has a lot more advanced features (e.g. incremental reading). But the tradeoffs are pretty significant, in my opinion. Anki's decisive features for me are its mobile apps, its simplicity, and its open source license. I've been using Anki for 5 years now and have about 10,000 notes. I don't use it for foreign language learning. I would not feel comfortable investing that much effort into a closed-source platf…

How are you able to keep 10000 notes in Anki organized? With that many notes, how can you tell whether something you want to remember is already in your notes or not? It seems to me that the Anki card browser is quite bad at this, but maybe I'm just bad at using it!

1. Anki has search, tags and hierarchical organization.

2. If you don't remember that you already have something, it means Anki is failing you, because the whole point of it is to help you remember.

3. BUT! it doesn't really matter if you add the same note several times, it will just make it appear for review more often.

Re: Effective learning: Rules of formulating knowledge (1999)

#28

Post by yowlingcat »

Am I the only one that gets nostalgic for university while reading this? Ah, I think back to university many years ago and it almost makes me a little sad. A time when I spent so much of my time learning that these techniques would come in handy. My brain felt so much more limber at that period. I am certainly faster and productive now at my craft, and am very happy with that. But, oh, to be a student again...

Re: Effective learning: Rules of formulating knowledge (1999)

#29

Post by iheartmemcache »

Eerie wrote: post #27
emilga wrote: post #24

Earlier quoted context omitted.

How are you able to keep 10000 notes in Anki organized? With that many notes, how can you tell whether something you want to remember is already in your notes or not? It seems to me that the Anki card browser is quite bad at this, but maybe I'm just bad at using it!

1. Anki has search, tags and hierarchical organization. 2. If you don't remember that you already have something, it means Anki is failing you, because the whole point of it is to help you remember. 3. BUT! it doesn't really matter if you add the same note several times, it will just make it appear for review more often.

It seems like the value of the platform is in the well-classified data made by the community in the same way that emacs "killer feature" is the bazillions of community .el extensions. Is there a Github or some defacto standardized way people publish their Anki notes?
Post reply on HN