I am quite cynic, but more repetition doesn’t necessarily move.
Especially when not somehow learnt once before.
Number of truly new items, even like 20, might still be too much; not to speak of 100.
However, there may be other ways to help, rather than waiting for repetition.
Extra study, ideally not just before you’re asked if you remember the meaning and reading of a kanji. How will you be sure you remember them?
We’ve had this discussion.
My direct experience with far more than just learning kanji is that repetition is the key to memorization.
Your brain apparently works differently, but physical flash cards and other repetition drills (like scales and arpeggios on a musical instrument) have been popular self-study methods for centuries, which makes me believe I’m not alone.
(To be clear, I’m not arguing against other types of study. Memorization doesn’t suffice to learn a language. But repetition trumps effort when it comes to memorization, in my experience.)
It’s funny, I had been thinking about this myself just a few weeks back since it was bugging me, but I figured it out.
In psychology courses, the idea of a “test” is often scrutinized, and the same can be said for sciences regarding the testing of hypothesis. So the question we ask is: “what makes a good test?” A good test is able to provide valid results to either support or refute the hypothesis. What do I mean by valid? Validity in this case means that the testing method is testing the target, and isn’t just testing irrelevant factors, and that by repeating the test, you are able to achieve a similar or same result.
So how does this apply to wanikani reviews?
To be honest, I don’t think the reviews are supposed to be a “test.” After all, that is why they are named reviews and not tests! A test would certainly be something like the Japanese Language Proficiency Test, which is a test meant to certify the level of proficiency with regard to the makers of the JLPT and their methodology. I think if wanikani had a “test” it would probably be something like, “every 5 levels, you must take a test for all the kanji in those 5 levels and pass with at least 80% accuracy- then you can pass to the next level.”
Because wanikani doesn’t have a feature like that, It is my understanding that the reviews are simply a means of upwards motion for the purpose of applying stakes to the “game” that is wanikani. It is as if you bought a really big deck of kanji flashcards and you start off with 10 flashcards, then you only let yourself add 10 more flashcards when you know you are ready for more.
So for a more direct opinion from me, I think that completing extra study before your review is completely okay. You are reviewing so many kanji and vocabulary that you will definitely have those kanji slip your mind enough for you to have genuine recall when you take the lesson. What I believe this is like, is just priming your brain for a more “high stakes” review. After all, the end result of “cheating” (not cheating, but this is to prove a point) in the case of doing extra study before the review, would be just getting MORE KANJI. And even then you will STILL need to review the kanji you “cheated” down the line.
TLDR: You’re just doing more practice. Which is good! There are no scores, the only measurement of your success with wanikani is your ability to consistently study everyday, even on the days you really don’t want to.
-Applied Linguistics nerd
I do extra study for recent mistakes two times a day, and whenever I do the kanji of a level (I always do all the kanji in one batch) then I do extra study: recent lessons every morning until I’ve gurued the kanji (while the kanji reviews come in the evening). I don’t think that is cheating because I can pretty much always remember all the kanji without any extra study when the guru2 reviews come along.
I wouldn’t do extra study right before doing reviews, but seeing items a bit more often when learning them or after failing them isn’t cheating yourself. The goal is to actually learn kanji so anything that helps you do that goes:)
Sorry if this has been said already… The very nature of extra study means you could never ‘cheat’ an item to burned because it won’t fit into the ‘recent lesson’ or ‘recent mistake’ category at some point, so I think it’s okay 
Though I do prefer to use extra study after a review session.
Well it’s not cheating but it is sabotaging the whole purpose of the SRS.
See the whole point of an SRS is to study each thing as little as possible and still learn it. With that efficiency, you can study more things with your fixed amount of time. In fact, you intentionally make the assumption that the SRS intervals will be enough to remind you frequently enough to remember forever (or a very long time at least), and trust the SRS to point out the small fraction of “hard” ones where that’s not the case with a failed review. You only study an item more when it’s proven that you need to. You’re letting the results tell YOU how much to study, not the other way around.
Now, if you want to study each item by writing it 100 times, you totally could do that. And you might even remember it better. But while you’re doing that, you’re not studying other things. SRS is a way to focus your study on things you know you need to study, and then more things, not waste time studying things “just in case”.
So yeah, you can study more than the minimum but all the ones you might have passed without it - remembered with one look - represent wasted time you could have spent on new items.
And if you’re already maxed out on volume, limited by the gated levels, then STILL don’t study current items more. Go use that time to read something or learn some grammar to reinforce what you do know.
Or, you know, go ahead and DO extra study. Everyone learns differently and you should do what works for you and not even feel bad about it. I’m just talking about the theoretical ideal way SRS is supposed to work.
THANK U GUYS! really, thank you very much to all of you who kindly commented on this topic. I appreciate it and will keep in mind all what you all have told me
The only problem is with an SRS it needs to be modified to each individual user. With WK it’s a set time deemed to be best for “most” based on the research Koichi had done before designing WK. Assuming he did the best research possible with it (which is doubtful based on the inherently flawed nature of the existing literature) at best he is hitting the perfect sweet spot for about 25% of the user base. For everyone else, their spacing is going to either be too long (roughly 25%) or too short (roughly 25%; where you would then have another roughly 12.5% where it’s way too long and 12.5% where it way too short).
For those who are lucky enough to be on the short end, they will feel really good about their study as they will rarely get anything wrong on their reviews. But of course, they are doing reviews more often than they really need to, which could be seen as wasting time, but I would argue that the boost to motivation makes up for that.
For those who are unlucky and they are too long, they will find that they are consistently failing numerous items in their reviews and may become rather discouraged with their study (rightly so). For this later group, the more exposure they have to apprentice items, the better. However, I would argue that doing the extra exposure right before the review, may not be enough to bridge the gap when it comes back as a guru item. If it’s a little too long, that extra exposure might be enough to lock it in, if not they will do pretty well with apprentice items, but rather crap with anything above. So it would be a safer bet to do the extra exposure after reviews so most of the items don’t just keep slingshotting back and forth between apprentice and guru.
So I would suggest if it keeps your motivation high go for it, but if you’re consistently failing guru items, try a new approach (i.e., due the extra study after reviews).
Yeah I seem to remember one of the versions of SuperMemo (the original SRS) kept personal statistics for both individual items and also for all items collectively, and slowly adjusted the interval spacing to keep you from getting 100% all the time or failing way too many. I can see why that would be way too much effort for a simple web app that works well enough for, I would say more like 80-90% of people. Perfect for hardly any, but not unusably bad for most.
I think Anki still does the individual items adjustment with the “ease” statistic, but I don’t think it keeps a general interval spacing fudge factor that’s applied to even future new items.
It would be interesting to have a slider in settings where you could choose to shorten or lengthen the interval spacings by like, 80-120%. Of course everyone would set it to the lowest at first because everyone thinks they’re special and need to go fast
This is a good reminder to not dismiss people doing extra studying. At the end of the day, how much studying and how many repetitions you need to learn an item on WK is up to you. You make that decision better than anyone else. The SRS timings might be set and work for many, but if you feel the need to study outside of that - just go ahead I feel. There is nothing inherently wrong in that.
But, as I mentioned above, I do believe it might not be the best use of your time. I don’t know how much time You have, I only recognize that many users are juggling life, work/studies and other things and so WK has to work within those restraints for a long period of time. Think about how to make it a pleasant journey, is how I see it. But what that entails, well, again, that’s up to each user. ^^
Yeah that’s about right. Based on the statistics I gave (once again assuming he actually chose the best spacing time, it would be good enough for 75-80% based on a normal distribution. One could say the type of people who would be choosing to study Japanese are already going to be above average compared to the general population, but the problem with that is that the research is mostly done on university students, who are already above average compared to the general population with things like memorization.
I personally like to do extra studies only for the first couple of days that new stuff enters my card pool, since there’s little to no spacing there and I haven’t developed enough repetition to ground the kanji/vocab into my head. The first few tries will very likely be 30-80% failures anyway, so why not try to hit the ground running and progress through WK’s time gates a bit more quickly?
FWIW, I’d say the whole point of an SRS is to let you get more frequent repetitions of the things you find difficult for whatever reason, and not waste time on the things you find easy or have already learned.
To me, it’s about letting you practice the hard stuff as MUCH as possible.
Hmm i don’t really get your question. Can you elaborate on that?
I think the question is: if you “quiz” yourself after six days, how will you know if you would have remembered it for seven days?
I just think people put too much importance on passing reviews. It’s flash cards, after all. You miss flash cards; it’s part of the learning process and doesn’t make you a bad person. I think that’s a product of the gamification - now I won’t level up! I’ve “set myself back”. Gamification can be motivating, but it can also tempt you to fudge the results.
Well yeah, which is a consequence of not spending your time on the easy ones un-necessarily.
You’re not quizzing yourself if you study right before the quiz. You’re always studying, which is good, but not the same.
jpdb.io has a setting for this (though they don’t quantify the effect, the options are just labelled “longer review intervals”, “even longer review intervals”, etc). The documentation says you can use it to trade off “how many reviews do I have to do” against “how many failed reviews do I get”. Personally I left it at the default value…
Shindii10,
I don’t know if you are aware, but there is a website called kaniwani.com that does the opposite of this website (shows English, and has you recall the Japanese word.
If you are truly concerned if you are actually learning the words on here, I highly suggest using it. My retention is significantly better with it.
EDIT: Depending on how committed you are, kamesame is far better, with a lot more features, but requires a japanese keyboard.