Yeah, I'm not sure how useful it is to use pure blanks with sentences like "boyfriend is [adjective]", particularly for katakana words.

A translation prompt would have the same pitfalls as straight vocab production cards. Reading-to-kanji blanks are good, but test something slightly different and don't need a paragraph of context.

I think gap exercises are useful while working on a text with new vocab. (What about sharing texts with the gap-filling, vocab or kanji quizzes they've made?) SRSing them months later, particularly if it requires remembering the exact content, seems less useful to me. Pure gap-filling SRS is good for testing expressions, though.

[edit: relaced my O/T SRS comment]

As for grammar gap filling, I wonder if JLPT-type exercises aren't better for understanding usage and differences? No risk of inadvertently making cards that have multiple answers or aren't actually a good example. And since there are so many JLPT exercises available, there's little need to create and SRS cards for grammar expressions. If they're already in SRS form, great - but I'd only repeat ones I got wrong. If JLPT1 includes some relatively expressions, just skip them.

MCDs can be easily created while reading a foreign text with "Learning With Texts (LWT)": you read, create a term (automatically with 1, 2, or 3 sentences of the text // "3" = "MCD mode"), and the system creates Anki MCD like facts, or you use the built-in SRS that supports MCDs, too.

Read more: http://lwt.sf.net

It's free, and in the public domain :-)
Enjoy.

Correct me if I'm wrong, but wasn't Khatz adamantly opposed to "fill in the blanks" style exercises? Even mocking it and saying that doing such exercises wouldn't help you in any practical way?

When I formulated the idea of taking passages of text from your reading materials and creating cards with multiple sentences per card, with some clozes on the front and single words on the back, it wasn't for learning new vocabulary words.

I've come up with various uses for clozes that I will further describe on the wiki in the future, but for learning new vocabulary words I think it's not such a good idea. I'm pretty much down even on sentence cards for learning new vocabulary, much less multiple sentences which at best strengthen the cue in the same modality, weakening it in terms of long-term effectiveness.

Even for single sentences now, I only bother if it's a single unknown word, and you have that word, unclozed in the sentence, as a cue on the front, enhanced/highlighted, and even then I make sure the entire bunsetsu is highlighted because it's not purely a card for learning new words but instead it's more of a bunsetsu card, a way of refining structural knowledge, and these cards are also more of a bonus card type since I have so many premade sentence cards in the decks I use as general reference corpora.

Edit: And especially if you're using the same sentence clusters/same sentences to learn multiple words across multiple cards, with the sentence(s) as a cue on the front you're just screwing up the intersession intervals by giving yourself the answers to the other cards, making the SRSing less effective.

Resurrecting this thread to ask a simple question that needs asking, that apparently no one has asked.

Does cloze deletion offer any improvement over normal sentence cards? Can anyone vouch for any additional benefit? Although it would seem like cloze deletion helps production, it may have no correlation whatsoever. It may just be making each card take longer due since you have to think about it. This lost time might mean MCD is not be as efficient as a faster, sentence based approach.

I cant find any studies of, or people praising cloze deletion other than saying it works. Well, so do normal flash cards. Using no flash cards works as well. The important thing here is to find out which is more efficient.

The polyglots which have blogs usually state that they learn with native materials, and mention flash cards can be useful, but none seem to use them extensively. I wonder if its even worth using cards at all after being able to tackle native stuff without much pain.

sikieiki Wrote:The polyglots which have blogs usually state that they learn with native materials, and mention flash cards can be useful, but none seem to use them extensively. I wonder if its even worth using cards at all after being able to tackle native stuff without much pain.

The reason they never mention using flashcards is because they don't anymore. Past a certain point their efficiency drops dramatically. You're spending more time creating and managing flashcards than it actually takes to learn the fact. You're also liable to go off on really weird unhelpful tangents as the kinds of things you learn become more and more granularly split.

This fascination with flashcards and systematic ways of studying all falls apart once you make it out of intermediate level. The advanced level of Japanese study just becomes reading, writing, conversation, and listening practice. Instead of testing for understanding of specific vocabulary you're broadening the scope to entire paragraphs, articles, short stories, or books. It just becomes about getting enough exposure to the language.

About the only thing I'm finding really useful now that I'm out of the intermediate level is the kanji flashcards I'm using to make sure I don't forget how to write kanji.

What erlog writes corresponds to my experience with English. After you reach certain level, and you are comfortable with it, you don't need SRS. Of course, there will be areas where your language skills are lacking but those areas tend to be specialist and it's a matter of whether you really want/need to acquire this specialist vocabulary. If you are a doctor, for example, and you want to work as such in your L2 country, you would need to acquire medicine specific vocabulary. At this point SRS might again become effective way of learning.