Hi guys.

Recently I've decided to learn perfect groups, singal primitives character readings with this deck https://ankiweb.net/shared/info/3283034296.

Adjusted it a little bit and made it to look like this:
http://imgur.com/4kP4wCj

I Put "signal primitive" and exemple characters on the front with the reading, exemples of words (a lot to add there, because it's not normally in the deck - have to add it manually) and exceptions.

Do you think such cards would be optimal for learning, would serve their purpose? Do you use any other, better decks that contain practical exemples of the words so that it will look like in a normal reading situation - not test-like?

The sorting in this deck is totally awful. Cards are sorted totally random. I want to learn the most important characters sets that have biggest readings coverage, so I had to sort the cards based on other sources, but I don't know If I did it right in the end.

Any suggestions?

(2015-12-26, 3:11 pm)Sztermel Wrote: Do you think such cards would be optimal for learning, would serve their purpose? Do you use any other, better decks that contain practical exemples of the words so that it will look like in a normal reading situation - not test-like?

Any suggestions?

It is easy to say that this is not optimal for learning, but not that easy to recommend something flawless.

1. I guess you are testing for へい and びょう. It is not good to test for more than one information per card. On the other hand you want to learn that the ON for 丙 is often へい and in case of 病院 or 病気 it is びょう.

2. I would recommend to use example words. You know the below deck which offers the RTK2 examples. You could test them one by one, then it is similar to learning vocab [Many people do recommend to learn just the vocab anyway]. But it is rather a MCD concentrating on the ON readings - so it would serve your purpose:
http://kazemakase.ca/remembering-the-kan...anki-deck/

Now let's look at your group cards: Minimum you need to bring in the question a hint that you need two answers for 病 by doubling 病. Combined with examples your question could look like this:

丙

丙 / 柄 / 病  /  病
丙: 甲乙丙 (こうおつ … )
柄: 横柄 (おう … )
病: 疾病 (しっ … )

病: 病院 ( … いん)

Better to make 2 cards out of this group question and test へい and びょう separately. That is easily done by highlighting the portion you want to test.

For the answer I would not use katakana and romaji, but of course once again the examples [But you just test へい or びょう and not e.g. whether you know 横柄]:

へい/ぺい  /  びょう

丙 => へい 甲乙丙 (こうおつへい) → A, B, C (old enumeration)
柄 => へい 横柄 (おうへい) → arrogance
病 => ぺい 疾病 (しっぺい) → illness

病 => びょう 病院 (びょういん) → hospital

Again you would highlight the portion you want to test.


You want to learn first the most important characters sets that have biggest readings coverage. RTK2 doesn't care about frequency it provides the complete RTK1 set including rare kanji and rare examples. Don't know how difficult your sorting was, but in the end it is a limited number of character sets you aim for. So if the sorting does require real efforts, just forget it and put this technical effort (matching different data sets) into getting examples over from your vocab deck or RTK2 or another available source.

BTW: The MCD is pre-made and therefore low effort, it could be an alternative for you - and you could also combine it with learning your group cards. As you seem to have frequency data available you could fiter out all rare examples (e.g. >30k).

Thanks for reply!

Now I have a dillema - switch to the deck provided on the website and start all over or modify my deck to be more recognition like, with exemples of the words on the front.
What I wanted to achieve was to learn the readings in anki with the deck (as I have no RTK2 or any book) and then easily recognize the characters and their readings in the wilderness or/as well have higher retention of the vocab that uses rare kanji - those words from vocab decks that use a kanji which is not used in other words, is rare.

And yes, I wanted to focus on the most important chracters - those who have the highest reading coverage, as I as well am strongly aware that learning readings through vocab learning is possible and perfectly fine.
Even though it's better, efficient to systematize some of this knowledge.

(2015-12-28, 8:27 am)Sztermel Wrote: Thanks for reply!

Now I have a dillema - switch to the deck provided on the website and start all over or modify my deck to be more recognition like, with exemples of the words on the front.
What I wanted to achieve was to learn the readings in anki with the deck (as I have no RTK2 or any book) and then easily recognize the characters and their readings in the wilderness or/as well have higher retention of the vocab that uses rare kanji - those words from vocab decks that use a kanji which is not used in other words, is rare.

And yes, I wanted to focus on the most important chracters - those who have the highest reading coverage, as I as well am strongly aware that learning readings through vocab learning is possible and perfectly fine.
Even though it's better, efficient to systematize some of this knowledge.

I had a look at the link for the deck you downloaded. The deck is from DrJones. It uses RTK2 info but also other sources like namakajiri plus findings from a thread on the forum.

So hopefully it is a good source with well defined groups, but surely it is impossible to match/combine it completely with the RTK2 deck. But you brought already in example characters - so I would assume you can get example words over too (not manually)*. That is even possible when you are already using the deck.

If you are already using the deck, you should know best, whether you need a general change of the setup or if just a few cards need changes (e.g. hints like doubling kanji, example words, or splitting).

It would make sense to first learn the ones with only one answer, then the ones with slight variants (セン / ゼン) until at the end the ones with multiple different readings plus exceptions.

DrJones goes with his groups beyond the pure, semi and mixed groups of RTK2, so it should contain quite some of the ones with multiple different readings plus exceptions. I assume you need to modify / split those to make them digestable - on the other hand DrJones and others have worked already successfully with it as it is. [When I first looked at your example card, I thought the answer was no good, but now it is clear, that the second row is coming from the exception field info. And there are also mnemonics for many of the more complex cards. So it might work out fine.]

In any case, I hope the deck contains also lots of unambigious/easy cards to start off!

----------------------------------

*In case you bring in example words, you might consider to set them apart or hide them. You could rate the difficulty according to whether the signal primitive alone triggers the answer, the example kanji are needed or if you needed also example words to get the answer.