who lol. You’re literally the only one who replied to my comment.
What information…All youve done is link a study that didnt say what you think it said.
Yeah and lets see what we have…basically everyone is saying that 1x1 is fine and its not gonna prevent you from being able to remember stuff…
Yeah, gotta be done wasting my time with this back and forth here lol.
With all due respect to Koichi, the interleaving of readings and meanings has never made any sense to me either. I want the meaning and reading to occupy the same place in my mind; that’s how reading works, right? I don’t think I’ve seen any evidence in this thread that Occam’s Razor fails in this scenario.
The problem with language learning sites – and this is by no means unique to WaniKani – is that folks will find a study without asking questions like:
On top of all of these issues, psychology is a budding field with very low replicability of results. See psychology’s replicability crisis or NPR on p-hacking.
At the same time, even when studies can be misleading, anecdotes are even worse.
As a language learner who wants to learn efficiently, it’s a really unfortunate situation. There’s so much misinformation, disinformation and suspect information floating around on the Internet about learning and it’s hard to parse out what actually might be useful.
Sorry for rambling, I don’t know that I’m going anywhere with this. These threads always end up seeming tragically futile to me. 仕方ない I guess.
I don’t know if I can really even call this an issue with WaniKani.
Putting the debate of interleaving to the side, the only way to maintain something after it is burned is to just keep reading Japanese material. The burned status isn’t there because you will somehow now retain it forever. The burned status is there because your review queue would become untenable if things floated around forever.
There is no kanji learning system that could teach you a large amount of kanji in such a way that retention of the material will reach a stage of complete and perfect retention. You will always have to keep reading Japanese literature to maintain what has been learned.
Ahh, not really, I wasn’t meaning to hone in on your example in particular, it’s more that this conversation as a whole made me think about studies and the issues with them.
I agree with everything you said. Research is better than nothing and anecdotes are also better than nothing. At the same time we have to take things with a grain of salt.
Are we still looking for sources on interleaved practice and why it works? I always feel sealioned when I participate in these discussions, but I’ll bite…
In this study, participants who used interleaved practice had lower practice scores but their scores on a subsequent test were doubled (this should sound familiar to anyone who’s had a day on WaniKani where they just couldn’t get one damn thing right).
This study showed superior results from mixed practice in the context of throwing beanbags.
This study showed superior results from mixed practice in the context of math problems.
This study also showed superior results from mixed practice in the context of math problems.
This study showed superior results from mixed practice in the context of anagram solving.
This study showed superior results from mixed practice in the context of identifying artists’ painting styles. Even though the respondents themselves insisted the massed practice led to better learning (because we all have a bias: when we think we’re learning and when we are learning are—and this is the point—two different things).
This study goes into the difference between momentary performance—knowing something because you heard about it recently—and long-term learning.
“[H]ybrid practice facilitated skill development more than blocked or interleaved practice alone. Furthermore, a dynamic interplay was detected among practice format, schedule, and learners’ prior knowledge.” If I’m reading it correctly, hybrid practice the least useful for going from input to output, but if the goal is to read and comprehend, you’re going to want to mix it up.
The idea of old kanji and new kanji overwriting each other is an example of what psychologists call negative transfer. Robert A. Bjork, Distinguished Professor of Psychology at UCLA, has devoted his entire career to studying memory and recall. There’s a whole book about Bjork’s research on the importance of forgetting some things (in this case, mnemonics, not kanji) to build space for others: Successful remembering and successful forgetting: A festschrift in honor of Robert A. Bjork. His wife Elizabeth is Professor of Psychology and Senior Vice Chair in the Psychology Department at UCLA and written a long list of papers on the same topic as well. If Tofugu did their homework, then I’m certain they encountered the Bjorks’ research along the way.
This is really cool. Thanks for looking into what I didn’t have the patience to do <3 I’ll have to take some time and read these. Thanks for the conversation, everyone!
Thanks! Several of these have the full text online somewhere or other too; some of the full-text results even show up in Google before the abstract results!
These are all interesting results, and to me they would suggest that kanji meaning/reading interleaving is worth looking into, but interleaving the meaning and the reading of a kanji seems potentially quite different from mixing up solving anagrams with throwing beanbags or interspersing math questions with history questions. I don’t feel that I can confidently stretch these results into our kanji learning context.
That said, I will consider taking breaks mid-WaniKani to play some cornhole or practice distinguishing Michelangelos from Rafaels.
Thank you all for the fruitful discussion.
Since writing this thread I have switched to interleaved lessons and reviews (aka. WK default mode.) Moreover, I’m only ordering my reviews (by SRS stage and then shuffled) on the days when I have a ton of Enlightened items piling up, in order not to leave lower stage items waiting. I’m trying to have as much as a “vanilla” WK experience as possible, even though I use an app.
It’s still early to say whether this will give me better retention than 1x1 mode, but in a few months my level 30+ items will start to get burned status, at which point I will be able to compare my performance between levels, say 25-29 (1x1 mode) and 30+ (interleaved).
As an aside, 1x1 always seemed to be a half-done, tacked-on thing, and disabling it makes other things work better. For example, when I get a card wrong, it will go away and reappear only after a couple minutes, as it should be, instead of being “stuck” on the front of the pile, which is something I never liked about the fake 1x1 mode provided by scripts.
As for language practice, I’m doing a lot of listening, but yes I should read more. Duly noted!
That’s what I would advise. You’re going to feel really frustrated at first, maybe for a while. But that’s around the level where I really started to feel my muscle-memory kick in when I’d do reading and listening practice. It won’t be much right now, but what you do see and understand will become so much easier to remember when you do your WaniKani reviews. It’ll feel like a lot less work.
If you have the extra energy, I’d recommend making a second SRS deck in Anki or japanese.io of non-WaniKani words. I only added words that I would’ve been ready for if they were in WaniKani. So at level 31, some useful words you could start with are 背負う, 騒動, 劇場版, 浮かぶ, 観覧車, 盗難… Find a book that you want to read and study words from it that you’re ready for.
Anyway, starting from the plainest version of the site and then adding features one at a time is a great way to go and more-or-less what I did as well. Good luck!
Go back through and create tables in Excel or Word of some of the older Kanji, print it out and reference it every week or so. Or just write them down on actual paper in a table format. Make your own flashcards, index cards are cheap. My Dad used to make fun of me, “Do you need more index cards from the dollar store?” I went through like 2 packs a month.
Nah. Just use Extra Study
I know about extra study, but if they’re having trouble with certain characters then physically writing them down in a different format would be helpful in building new mental connections. It’s more work, but doing more work gives greater exposure. They could also write their own sentences out of words they don’t see much.
It’s been 6 months and I’m still at level 30. I haven’t been able to advance a single level, because:
whenever an older item (Guru / Enlightened) comes up for review, I remember absolutely nothing about it and so it fails several times down to Apprentice, at which point I have to slowly try to get it back into memory;
during review I investigate all the “burned” kanji and radicals that I forgot and have to Resurrect them all
this makes my review pile to be constantly growing, instead of shrinking, up to hundreds and now thousands
Therefore I decided to stop doing newer lessons until I see a zero in my review stack (which won’t happen anytime soon.) Fortunately I originally bought a lifetime account, so this is not costing me money.
But that still wasn’t enough, so I decided to follow this advice: How to kill a giant review backlog
I put my WK in vacation mode and started doing custom studies from level 1, one level at a time, for as long as it will take.
At the same time, I’m making my own Anki deck to practice WRITING all the kanji, without using mnemonics, because I understood that my trouble with similar kanji stems from the fact that I haven’t been writing them at all since I left language school.
So for every kanji, from level 1 onwards, I find the simplest or most common word whose reading is unique in the dictionary and I use that as my Anki question.
For example, this question:
Only has one possible answer:
(I know it’s 花火, but this is the card for 花 so everything else gets the hiragana treatment.)
I chose はなび because はな can be 鼻, 華, 洟… What I do is write 花 with my finger on my phone’s screen (it’s an AnkiDroid feature called whiteboard) then I flip the card and rate myself.
It’s still early to say if and how much it will help, but I’m seeing some progress.
I will share the deck when I get back to level 30, in case others want to use it.
To anybody who’s wondering how is it possible to get to level 30 before “discovering” you’re forgetting most of your burned items: I suppose my short + medium term memory is good, so much so that I can get items to Guru using just that.
But I’m missing the bridge to long term memory, therefore sooner or later they all invariably get vacuum cleaned from my brain. Or rather, my long term memory was (is?) still based on what I learned in Japan, with all the emotions and stuff attached to it, and my WK use was not contributing to it at all.
I could power through and learn level 31 in a week at any time if I wanted, and then 32, 33… I know I could, because I’ve done it for all previous levels. But what’s the point, when you’re forgetting how to read and write 花?
Are you interacting with any written Japanese content? I can see not remembering a kanji if you only see it literally once every month or so, but that’s not what you’re supposed to do.
I have forgotten burned items, but only the uncommon ones. And I’m also not great at visualizing, so I struggle with remembering them enough for writing, but that’s not what I’m looking for, so it’s not a problem for me.
If you want to actually write them, you do need to practice writing them a lot. Not once every time they come up, but 10 times every day until they stick.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.