RECENT TOPICS » View all
Hi everyone
I know a similar question was asked several times about using both systems and the answers have mostly been that it's a waste of time.
About that, I don't care. But what's bothering me is wether using both is bad or could even harm your stuy. I thought about this because a SRS has its reviewing time on purpose and if you're doing more than one at the same time, you're going kinda against that. (I hope it's understandable what I mean ^^)
At the moment I'm doing this: Learning a Kanji from RTK1 and write it down. Then, after I'm through the amount I wanted, I add them here and review it. After that, I add the Kanjis in Anki and review them the next day.
I've also one further quesion which is not really related to RTK (I hope it's fine if I ask anyway). How do you learn new vocabulary? When doing other stuff besides RTK (grammar i.e.) you automatically read new words.
So do you guys learn the Kanji -> reading -> translation or reading -> translation or
Kanji -> translation?
Thank you very much for your answer.
Using multiple SRS can harm your ability to store the facts in long-term memory. You'll still manage it eventually, but it goes against the core principle of SR - showing your brain that cards keep showing up, even after long breaks.
So two SRS programs working in parallel will slow you down, and will take more time to review on. Just take one and spend that time to just do more cards or use other methods instead.
New vocab is first learned through core6k, as it grants you the ability to READ kanji in context (I know about 40% of the core6k words I've come across, but can't read them). After you can read real text, you just move on to..reading real text, which is where the rest of your dictionary will be formed from.
Fadeway wrote:
Using multiple SRS can harm your ability to store the facts in long-term memory.
Where did you learn this? I've heard it said a lot but I've never seen any evidence to support it. If you know of any studies that look in to I'd like to know because it seems to me that this belief has been perpetuated by people stating it as a fact and causing others to believe and repeat it.
This isn't an attack on you, I'd genuinely like to know what the basis is for this, but it seems odd to believe that repeating something will lessen your ability to remember it. Has anyone even studied that? I thought the point of SRSs was to let you spend the minimum amount of time on each fact (thus allowing you to learn large volumes of information), not to commit each item to long term memory as quickly as possible.
If that was the case everytime you opened a book would screw up the spacong.
Multiple reviewing does not "harm" or deter long term memory potentiation. The reason that blackbrich alone gave would prove that plus the rote-memorization we all did and people around the world do.
The bit that's being confused and twisted to mean something else here is this. When we learn something/store it in memory, you can think of it as setting a timer on that "chunk of knowledge." When that timer runs out you forget it. Reviewing that knowledge at different points on the timer's time will help refresh the timer. The most OPTIMAL execution is to try and review a chunk of knowledge just as the timer is going off or right after it has gone off. Its optimal like this because you can spend the rest of the time reviewing/learning new stuff instead of constantly refreshing the timer on stuff you don't need to. One of the other things that I understand is that if you wait for the timer to go past a certain point before reviewing again, you will actually grow the new timer. So maybe the old timer was 10 days till forget, and now you reviewed in the 'sweet spot' so the next time till forget is 20 days. You are always constantly trying to hit that sweet spot because as we all known from school days, constantly re-reading something doesn't always mean it sticks long term. (EDIT: I guess you can think of the sweet spot as when we have moved something out of Short term and into some stage of long term, while its stuck in short term you won't be doing much in the way of long term potentiation).
This "timer" analogy is known as the forgetting curve and is what SuperMemo is applying (and Anki is based off of a SuperMemo algorithm). Try reading some of the articles over at the SuperMemo site, they are informative.
Last edited by vix86 (2012 April 04, 1:23 pm)
I'm being entirely speculative, so if you only care about scientific evidence, you can safely ignore this post.
You're both correct. Using SRS is meant to help you remember with the least amount of repetition. Heavy usage, like reading, is what we all strive for, but when learning kanji from RTK1, you can't really read. Back when I was taking new kanji in RTK, when I encountered IRL (or, more commonly, thought of the shape of) a kanji I was supposed to know the meaning of but didn't, I'd often struggle for half an hour during class to remember, and when I finally did, it'd often turn out to be one of those kanji that have just entered mature intervals. When a week later that kanji showed up on SRS, I'd get it right, but would be reluctant to send it to the 3-month interval, since I had only gotten it right because of having seen it a week prior, and would either mark it wrong or hard - I knew I'd have no chance of getting it right if I sent it on a 3-month interval.
If I were to remember the kanji easily, instead of struggling, I'd have involuntarily decreased the interval Anki meant me to take (say, 28 days) to a smaller one (say, 21 days), making the next interval potentially harder and increasing fail rate for it(and next intervals are always twice or thrice costier when failed, when it comes to time).
Doing two SRS programs at a time will reproduce this effect, except it'll be happening always, not once in a while. And do the math: if Anki intervals are 1-4-15-28-90 (just making up numbers), on a normal schedule, you'd get to 90-day retention within 138 days, assuming you always answer "good". If you do double SRS, you'd be getting two reviews for the 90-day period, so you'll only have a 45 - day retention, and the intervals will not be steadily larger, due to interference. Thus, to get 90-day retention, you'd need to get to a 3-month interval on one program, which may or may not take longer, depending on how often you make mistakes with single SRS. In any case, it's better to just take the new Anki and adjust the algorithm accordingly, it'll avoid the interference at least.
PS: Great post vix.
Last edited by Fadeway (2012 April 04, 1:51 pm)
Thank you very much for these detailed answers. So two SRS programs wont' harm, that's good. But it isn't an advantage either. Then I'll stick only with this page for the moment and see how it is going.
Again, thank you
Thanks for the links Vix.

