I've been having trouble remembering cards after learning them in the study area and by the time they've expired I've already forgotten the story.

So recently I've been testing the cards I've just recently learned about 15 mins to half an hour after I've learnt them. It seems to work and so what I've been doing is every time I log in I test myself on the expired cards and stack 1 as well.

The result is that I don't have any cards in the first stack at all now, and I want to know, is this counter-productive to my studying? Because, after all, I'm not using the spaced repetition system exactly as it was designed.

Short Answer: Don't review from green boxes, long answer follows:

It's very counter-productive, your reviewing stuff that's stored in short term memory, the entire goal of the Leitner system is to slowly but gradually move information from short-term memory to long-term memory.

You're only cheating yourself, only review from failed or expired boxes.

This seems like a good place to bring up a similar problem I've been facing recently:

If you add new cards, you can review them in 24 hours and then 3 days later from the first box. If you fail a kanji and reformulate it, the earliest you'll see it is 3 days later from the first box. A lot of times I find my new (reformulated) kanji has been forgotten after 3 days. In the end it goes right back to the failed kanji stack.

Has anyone figured out how to get an extra review after 24 hours? I don't actually want to do the first box reviews early, for the reasons meolox mentioned.

unzum Wrote:I've been having trouble remembering cards after learning them in the study area and by the time they've expired I've already forgotten the story.

After you have created a story, wait at least until the next day before you click on the blue stack. If you can`t remember the kanji after one day, you know you need to work on the story more.
When I fail a kanji, I also wait at least one day and do a mini quiz before I hit "learned" on the study page. Does that make sense?
Then you shouldn`t have as much trouble remembering them before they expire. (But as meolox said, don`t test the green cards)

scout Wrote:Has anyone figured out how to get an extra review after 24 hours? I don't actually want to do the first box reviews early, for the reasons meolox mentioned.

There are a couple threads somewhere on how to review the failed cards the same as you review the expired cards.
But what you can do in the study page, is look at the next keyword and draw the kanji for it before you hit learned on the current kanji you are on. When you hit "learned" that next kanji will come up, if you got it right, quiz yourself on the next one before you hit learned, and continue on down the list. If you get one wrong, just leave it in the failed stack (don`t hit learned) and work on the visualizing the story more.
Does that answer your question?

EDIT: from this thread:

synewave Wrote:You can review off the red stack if you use:

http://www.kanji.koohii.com/review/revie...ed=1&box=1

@laner36
I've seen the threads about reviewing the failed cards as a stack, but that only seems useful for weeding out failed kanji that you aren't really failing at the moment. (i.e. you failed them during a prior review, but for whatever reason you can recall them right now.)

I've been letting the failed stack pile up as I push through the rest of the book. I do a decent amount of reading on the side, and I occasionally run into characters that are in the failed stack. This tends to give me enough of an incentive to get them out of the failed stack. Usually I get one or two this way every few days. I could write myself a note to go back at look at them in 24 hours, but it really seems like something that the SRS should be handling automatically for me.

scout Wrote:I've seen the threads about reviewing the failed cards as a stack, but that only seems useful for weeding out failed kanji that you aren't really failing at the moment. (i.e. you failed them during a prior review, but for whatever reason you can recall them right now.)

The ones you don`t "weed out", you know you need to spend more time visualizing the story. After reading various posts from Fabrice, (I could easily be wrong, but) I think the reason he doesn`t have a 24 hour review stack, is to encourage us to really visualize the story. So if you can`t weed out a kanji after 24hours, then you should look at your story and see why it is not sticking.
If you want to drill drill drill, then an SRS like anki might be good for you. But after trying both, i would recommend going with this site for Heisig kanji--well i would go with working on your story and your creative memorization rather than just memorizing by repition no matter what SRS you use.

scout Wrote:I've been letting the failed stack pile up as I push through the rest of the book.

One more thing: I don`t know how far you are in the book, but unless you are near the end, I would recommend you spend some time on working down this failed stack. A lot of kanji you learn early in the book comes back as primitives later in the book (as you have probably already seen).

Yeah, try to keep the failed stack to a minimum, and weed it out every day. I keep mine under 20 or so now.

I have a strict policy on letting things out of the failed stack-- I only release it after I can recall it from keyword->kanji successfully. So on any given day, I go through my failed stack of yesterday's stuff, see which keywords I can get the kanji from, mark those learned, then go back and study anything I can't pass that day. I've had some words sit in that pile for a couple of days-- 4 at the most-- but I'll fail anything I don't get right, and if I fail it, I study it, then don't touch it for a day. Then I do the day's reviews, and only after that do I add any new stuff.

I'm also tracking things that show up in the fail pile a lot. There are some kanji that just don't seem to stick. I've tried all sorts of things to remember them, so I'm just going to fall back on brute force repetition for those few.

laner36 Wrote:One more thing: I don`t know how far you are in the book, but unless you are near the end, I would recommend you spend some time on working down this failed stack.

I appreciate your input and I guess a bit of background could be helpful. I've gone through the entire book and created stories, however I only have ~1500 entered into the SRS. Around the time I got to ~1000 a coworker was reading several papers by the author of SuperMemo who apparently advocated only adding things to an SRS that you truly "knew". So I basically held off on adding new cards until I'd looked at them several times. The rate I was adding cards slipped as I realized that for the amount of time I had each day, I would either do a good job creating stories, or a good job of reviewing new kanji. I found I was good for creating stories in the morning, but then wasn't necessarily feeling good enough to do reviews later. I had a really hard time finding enough time to do both well, so I effectively split the process. I now have my own stories entered for all of the kanji, and I am going through and slowly adding the rest in.

laner36 Wrote:The ones you don`t "weed out", you know you need to spend more time visualizing the story. After reading various posts from Fabrice, (I could easily be wrong, but) I think the reason he doesn`t have a 24 hour review stack, is to encourage us to really visualize the story. So if you can`t weed out a kanji after 24hours, then you should look at your story and see why it is not sticking.

My failed stack is at ~200, so going through the whole thing when I've only reformulated 1 or 2 seems like a huge waste of time. I really don't see why a reformulated card should be handled any differently from a newly added card (i.e. it should go into the blue stack.)

Before, I very strongly felt that failed cards should go to the blue stack before going to the first stack. But nowadays I don`t think it is a big deal. Perhaps because I figured Fabrice wouldn`t change his mind about it (so I tried to implement it into my own method). Or maybe I understand better now why Fabrice doesn`t do it.
Anyways scout, Your method of adding cards sounds very different from mine so I don`t really know what to tell you. But I would never go through 200 failed cards--as you say, that would be a waste of time.
Let me try to explain better how I do it (or did it) and hope that it helps somehow...

[Continue reading for a (boring) description of my failed stack method...]

For me, a majority of cards that end up in the failed stack don`t need their story "reformulated". Kanji are failed for many reasons (eg. being tired, missing a stroke, or the story being on the tip of the tongue) but it doesn`t necessarily mean that the story needs to be changed.
I try to look at the kanji right after it is failed and I know if I need to change the story (because I have missed this kanji 5 times, etc) or if I can just quiz it the next day and get it back in the system.
When I had a huge pile of failed cards, I would add new cards until I had my failed stack to under about 20. (Also I never used the link synewave gave, I would go through them the way I tried to explain above, clicking learned on the kanji I remembered perfectly the day after I failed them the first time). I would stop after there were 10 that I could not remember after 1 day. These 10 i would look at again and at _that_ point reformulate their stories if needs be. Then the next day I would usually be able to get through these 10 and "weed out" a few more until I again had 10 that I couldn`t remember from the day before.
So if I had 200 in my failed stack, I would push "learned" on maybe 30 kanji before I had 10 in the study page that I couldn`t remember. than the next day (**before reviewing expired cards**) I would go through the 9 or so from the day before and maybe 30 more before having 10 on my study page. That day I would fail some more from the expired cards but my failed stack would be down to about 140 or so. I would continue in this fashion until my failed cards were down to 20-30 before I would start adding more cards (as new cards are generally failed more often).

EDIT: In other words, the only reason I would go through all 200 failed cards in one day is if I hit "learned" on 190 or more of them (or "weeded them out").

@laner36
Thanks for the detailed reply. If I'm understanding your system right, it worked because you were going through the failed stack linearly, so you could re-review the same subset of cards the next day. Is that right?

In general I'd agree that going through failed cards sounds like a wiser use of time than adding new cards, but I haven't found that to be the case right now. It seems that ~90% of the new cards I add make it to the 4th stack. The ones that I fail I tend to fail pretty "hard". Either the story just doesn't seem to stick well, or I'm constantly confusing it with another keyboard, etc.

I'm sure I'll see better in hindsight how I should have gone about this, but at least at the moment adding new kanji seems to be the most effective way of increasing the total number of kanji that I'm proficient at.

Scout--have you thought about just reviewing 20-30 from your failed stack each day? That way you could knock the whole thing out a couple of weeks. Just go down the list of failed kanji (don't go into review mode-- just look at the long list you get when you hit "View All"), and just focus on 20-30 of those a day. Since the default list goes from low kanji number to high numbers, the ones near the top should be ones you have a better shot of remembering. Even if you can't clear 20-30 a day, you should be in a better position to clear them the next day. I've had as many as 80-90 in my fail pile at one time, and I managed to plow through it.

Once you have it cleared, it's a LOT easier to keep clear, and that way you don't ignore ones you have trouble with. Because you're going to reach a point where your failed pile has 250+ in it, and you'll be stuck there until you clear it. Then you'll have to get them through all those reviews, and that can become a big drag on your momentum. (And the last thing you want is that.)

I wouldn't worry so much about reworking stories-- sometimes nothing will stick no matter what you do. You just have to keep seeing it until your brain figures out that it's important to remember it. I'd only rework a story if I have a low pass/fail ratio, or just an inordinate number of fails.

Thanks for the replies everyone.

It's all pretty much sorted out now. My failed stack is down to 0 (yay!) so I won't need to test from the green stack anymore. The reason my stories weren't sticking was because last year I started Heisig and did 600 kanji before stopping, but because I didn't have this site back then I couldn't remember almost half of kanji when I started testing again. It wasn't little problems with remembering these kanji, I needed to make up completely new stories for them.

But now I'm back to learning new cards and I've improved my mnemonic/visualization skills so I should be fine.