I was wondering what your thoughts are on this.

Pareto's Principle says that 20% of the things we are doing gives us 80% of our results.
Vice versa, 80% of our efforts only brings 20% of our results.

The implication here is to drop the 80% we are doing right now, and focus more of our time on the methods we are only using 20% of the time.

So in your opintion, what would these 20% be? Maybe be as precise as you can.

Some people say learning through reading is much more effective with japanese if you know Kanji, because you understand the parts of the new vocab you are learning. And for learning it is much more effective than listening.

I would say for me that most effective has been using ANKI regularly, every day, and with simple material that can be reviewed fast.
I've also done other things (text books, lessons, immersion with native material, Jpod101...), but I think simple material reviewed in Anki regularly was probably the most effective.

What would be your top 20% of your efforts, which are more effective than others?

Seems related to the Zipf's law.

What 20% you focus on will probably depend on where you are in your studies. Maybe in the beginning that would be getting down the basic grammar. From Intermediate to Adv though, I think the best 20% could be spent cramming vocabulary; the payoffs are huge.

The Pareto Principle applies to some systems, not all. I disagree that it applies to language learning. Language learning comes down to time spent receiving comprehensible input...the specific methods are less important than just spending the time on it.

Yeah, I'm not sure it's all that useful a concept for choosing studying techniques.

It does find it's way into language learning though--since the most common grammar or vocabulary appears very very often. It be pretty close to a 20:80 ratio.

There was a post recently in the Koohii Lounge/Off topic part of the forum about this very thing

http://forum.koohii.com/showthread.php?tid=13022

The video referenced went over this and was quite enlightening.

Yes, the video referenced in that thread is a great explanation.

For language, Zipf's law and the Pareto Principle only seem to apply to vocabulary; the most common words are the majority of the words on a page, and yet you can probably count them on your hands.

This is great for when your first learning, because you instantly see results if you're learning vocab based on frequency. Unfortunately it's also the cause of the plateau.
Of course, you're not going to fail to understand 100% of a book because you don't know every word in a language, but it will take significantly longer for that to be true of any random (contemporary) book than to understand 80% of the same. And since proper understanding is claimed to be achieved above 90%, that's quite a bit of extra work for just 10+%

As for studying, this is why I like frequency lists: minimal effort, maximum effect.

vix86,
I agree, well for me I am a lower intermediate at this point. I am in fact focussed on vocab right now, the question I have is more related to, which methods are the ones that are most effective for that.
"depends on the person and situation" and so on is of course always true, but I'd be interested in personal experience.
Which 20% of the methods were highly effective and which weren't?

Standfield123,
I spent quite some time on things that just don't work as well as others. For instance high level immersion may be not nearly as effective as immersion on your level (like you say comprehsensible input). Textbooks are comprehensive input, but might be not as effective as working through graded readers, for instance, and so on.
That's what I mean.

Tzadeck,
Fair enough, however I don't really see why? in my experience not all teaching methods I tried have had the same level of effectiveness. Have you really never tried a certain studying technique and then thought that it wasn't for you? Then it would belong in the 80% group for the sake of this argument.

sholum,
yeah that makes sense and relates strongly to what I mean. Is there an ANKI vocab deck that is based on frequency? I am not sure about Nayrs and Core2k in that regard?. The optimized Core2k while obviously being a usful deck includes words like 総理大臣. How important is that going to be in everyday conversations?

maxwell777 Wrote:Is there an ANKI vocab deck that is based on frequency? I am not sure about Nayrs and Core2k in that regard?. The optimized Core2k while obviously being a usful deck includes words like 総理大臣. How important is that going to be in everyday conversations?

Both those decks are frequncy based. But what you choose for your input for frequency also matters a lot. The core decks are as far as I know based on newspapers and with regards to newspapers 総理大臣 might be an important word. What 'daily conversation' covers is difficult to say I guess. It might somewhat on who you are and there are just a lot of topics you can easily cover.

Out of all the unique words in the corpus, half (approx.) occur only one time.
In my novel text of 40,000 words, there were 4,000 different words. 2,000 of them occurred only one time in the text. I guess that's not too bad. 2,000 words (that make 95% of the text) is still quite a lot.

maxwell777 Wrote:vix86,
I agree, well for me I am a lower intermediate at this point. I am in fact focussed on vocab right now, the question I have is more related to, which methods are the ones that are most effective for that.
"depends on the person and situation" and so on is of course always true, but I'd be interested in personal experience.
Which 20% of the methods were highly effective and which weren't?

yeah that makes sense and relates strongly to what I mean. Is there an ANKI vocab deck that is based on frequency? I am not sure about Nayrs and Core2k in that regard?. The optimized Core2k while obviously being a usful deck includes words like 総理大臣. How important is that going to be in everyday conversations?

Method only matters in what you want to focus and what you find easier to work with. For example, if you want reading only then obviously you'll focus only recognition, but if speech is important, maybe you just focus listening comprehension.

As far as decks go, there is no "defacto" deck. Even Core2k is based off a frequency list. It all comes down to the corpus that was used to build it. But whether the words are "useful" or not is meaningless point in my opinion, above a certain word count. At a certain point, you'll be learning words that only have 1-9% occurance if not less. One of the major points of Zipfs law is that the frequency of word usage drops off rather quickly.

[Zgarbas]

Quote:The implication here is to drop the 80% we are doing right now, and focus more of our time on the methods we are only using 20% of the time.

I think that that's a bit of misunderstanding of the 80/20 principle. Even if you switch entirely to a more efficient method, you won't achieve 100% efficiency - rather, you will still have those things that you struggle with which will naturally take up more time. It's not about efficiency regarding study methods so much as the inevitability that individual factors will weigh in differently. For example, say you are studying 100 words, and for some reason 20 do not stick (a fair percentage) - you will dedicate more time to those 20. Unfortunately, the arbitrary ease regarding what sticks and what doesn't will be a constant regardless of your method (though of course, a more efficient method will decrease your overall time investment and enhance your results). And as previously mentioned, chances are that the first 1000 words you learn will see far more usage than the following 5000.

I'd say I use pretty much the same basic vocabulary 80% of the time or more... but since I like talking about a variety of topics I end up using the not-as-useful words every now and then. While not as significant, usage-wise, I find them important as they enhance my overall conversation ability and, well, make me sound like a conversation partner rather than someone who is struggling too much with the language to actually talk about things. I think that even the most obscure I'll-never-use-this-word-why-on-Earth-am-I-memorizing-the-dictionary words have popped up in my life at least once, somehow*; is that one time worth the investment? It depends on your personal goals and needs.

*I am fairly sure that no one would ever expect me to know some of those words, judging by the crazy look people would sometimes give me afterwards...

maxwell777 Wrote:I was wondering what your thoughts are on this.

Pareto's Principle says that 20% of the things we are doing gives us 80% of our results.
Vice versa, 80% of our efforts only brings 20% of our results.

The implication here is to drop the 80% we are doing right now, and focus more of our time on the methods we are only using 20% of the time.

So in your opintion, what would these 20% be? Maybe be as precise as you can.

Some people say learning through reading is much more effective with japanese if you know Kanji, because you understand the parts of the new vocab you are learning. And for learning it is much more effective than listening.

I would say for me that most effective has been using ANKI regularly, every day, and with simple material that can be reviewed fast.
I've also done other things (text books, lessons, immersion with native material, Jpod101...), but I think simple material reviewed in Anki regularly was probably the most effective.

What would be your top 20% of your efforts, which are more effective than others?

I like this question. For me the "20% of things that give me 80% of the benefit" are:

1. Using anki daily to help build up my language knowledge (kanji, vocab, grammar)
2. Weekly lessons with a professional, native speaking teacher
3. Taking the JLPT annually

For me the pareto principle hasn't meant "dropping" the other 80% of activities. Rather, it meant really putting a lot of effort into optimizing these 3 activities.

An example of me optimizing anki was learning how to create my own grammar decks and changing my kanji deck to be recognition and kanji->definition.

An example of optimizing weekly lessons was moving from in-person lessons to lessons via JOI.

By taking the JLPT annually I get professional, objective benchmarks for my language knowledge, reading comprension and listening comprehension abilities, and know how close I am to hitting them.

The 80% that I've depriotized is the fun stuff, which for me is speaking with native speakers.

I know a lot of people, both in Japan and out, who never break out of the beginner level. By focusing on these things I moved out of the beginner level and into the lower intermediate level.

Take 2 standard Novels/Visual Novels in genre that you love. Assume that you study every vocabulary and grammar pattern that appears. You will be able to understand 80% of all (written) Japanese media (not classical) and 90% of media in your genre. Here how's it works out:
-80% of the media is made up of 20% words in the language. We can say these are the "common" words.
-80% of the text in first novel would contain words from the 20% pool. Same for second novel.
-So essentially you are learning 80% of the words from 20% pool, and 20% words from 80% pool for each novel.
-Easier the novel, the greater your chances of hitting the 20% pool.
This is although impractical (and I'd never to advise anyone, especially beginners), works.