I just recently passed the halfway mark of the first volume. Here are some thoughts I have on Heisig's method, especially using this website:
Keywords
As I am not a native speaker, I often don't know Heisig's keywords. That isn't so bad actually, as I tend to associate kanji with concepts rather than a single word. However, there's an annoyingly large number of synonyms, which become a problem when you're not so confident in English. I can overcome this by incorporating clues into my stories to distinguish synonyms (usually relying on wordplays). The reward for this additional work is the acquisition of some new English vocabulary.
However, I'm having doubts that English is such a good source language to be basing keywords on, mainly due to its high ambiguity. If your native tongue seems more suited, you should carefully consider what language to use.
Stories
Having read a few thousand stories on this website, I adapted what I consider bad habits:
Just using a single sentence, slogan or rhyme. Even if they sound compelling at the moment you make up your mnemonic, for me they tend not to work.
Better than a sentence, is always an image. And I really have to force myself to visualize it.
Describing this image in form of text, often sounds very arbitrary and incomprehensive. But our - or at least my - brain works different. What you remember is not the words you are reading, but the image that is created in your mind. So please don't hesitate to share 'stories' that sound rather boring.
If a single image doesn't suffice, it always helped me to turn it into a more complex story (with actions).
The "Spaced Repetition Sieve"
A technique I subconciously started using, is sieving out difficult kanji through this website. I went through the list really quick, only thinking about a single image, or sentence for all kanji. I then added a whole lot of new frames at once, and went through those the next day. For the first repetition, I then fail maybe 80% of those. But those 20% that pass, are sieved out: For those, I won't need a complicated story.
For the kanji that I don't remember, I create a more complex image. If it still fails, I will eventually turn it into an actual story. By this process of iteration, I use the right amount of complexity for each kanji.. or at least that is my theory. Does anyone else use a similar method? Any objections?
Keywords
As I am not a native speaker, I often don't know Heisig's keywords. That isn't so bad actually, as I tend to associate kanji with concepts rather than a single word. However, there's an annoyingly large number of synonyms, which become a problem when you're not so confident in English. I can overcome this by incorporating clues into my stories to distinguish synonyms (usually relying on wordplays). The reward for this additional work is the acquisition of some new English vocabulary.
However, I'm having doubts that English is such a good source language to be basing keywords on, mainly due to its high ambiguity. If your native tongue seems more suited, you should carefully consider what language to use.
Stories
Having read a few thousand stories on this website, I adapted what I consider bad habits:
Just using a single sentence, slogan or rhyme. Even if they sound compelling at the moment you make up your mnemonic, for me they tend not to work.
Better than a sentence, is always an image. And I really have to force myself to visualize it.
Describing this image in form of text, often sounds very arbitrary and incomprehensive. But our - or at least my - brain works different. What you remember is not the words you are reading, but the image that is created in your mind. So please don't hesitate to share 'stories' that sound rather boring.
If a single image doesn't suffice, it always helped me to turn it into a more complex story (with actions).
The "Spaced Repetition Sieve"
A technique I subconciously started using, is sieving out difficult kanji through this website. I went through the list really quick, only thinking about a single image, or sentence for all kanji. I then added a whole lot of new frames at once, and went through those the next day. For the first repetition, I then fail maybe 80% of those. But those 20% that pass, are sieved out: For those, I won't need a complicated story.
For the kanji that I don't remember, I create a more complex image. If it still fails, I will eventually turn it into an actual story. By this process of iteration, I use the right amount of complexity for each kanji.. or at least that is my theory. Does anyone else use a similar method? Any objections?
Edited: 2010-10-25, 4:15 am

. I don't put them in RevTK until I've managed them once with the flashcards.