Language Journal: August 14th 2019
This is more a post to talk about yet another change to my study process.
I've largely started abandoning all buttons except Again and Great in Anki.
This flies in the face of most recommendations (which recommend avoiding the Great button). But, I think it actually makes perfect sense considering all things like my objectives, number of reviews and so forth.
As I said in the past... unless you're studying for a specific test and need accuracy on a smaller scale, faster, then you actually want less than 100% accuracy. If you're getting perfect scores, that means you're studying at an EXACTLY perfect rate (highly unlikely), or that you're babying yourself (MUCH more likely).
Furthermore, when you're getting perfect scores, it is impossible to measure improvements in your methodology and even harder to convince yourself to seek improvements.
My review load isn't massive. But it is also not small. On a typical day, I have between 150-210 reviews at the moment. Usually somewhere in the middle of that. And it takes me 35-45 minutes. I typically get an accuracy of 95%. Not perfect. But, close.
For me, this is too close to perfect for my liking. I'm still complacent with the results. So, I was thinking about how to "fix" the situation. And the solution I came up with is somewhat counter intuitive.
By preferring buttons which push my cards off for shorter periods of time, I'm maintaining a consistent accuracy and review load, primarily by balancing out babying myself, with the more frequent exposure. In short, there is a balance, and it comes without any cost to me. It is almost a self-maintaining balance. The cards are more frequent, but that also makes them easier to remember, which decreases my odds of hitting "again".
To break this, by preferring Again and the maximum interval, I'm pretty much always punishing myself on the cards I haven't mastered, while properly kicking out the ones I have mastered. I'm a bit more flexible with newer cards (21 days or less). But, once they are out past that, if I get them right, they get kicked far down the road, and if I fail anything on the card, it is back to basics.
If I REALLY do know my cards. Then the review load will drop. If I've been babying myself though, the load will increase and the only solution will be to find better ways to commit the cards I'm failing to memory. And, as I improve my methods for remembering, the load will drop again and faster than it would if I stuck with shorter intervals.
Basically, I inverted the recommendations for the sake of putting the onus for improving on me.
The suspicion here is that at the moment I'm being bombarded by a mix of cards I don't know well enough and ones I know too well. And that isn't a problem in and of itself. But, the ratio of cards I know too well, is too high. And of course that the current rates of change are too forgiving and not putting enough pressure on me.
If I'm right, then after a few weeks my daily review load might drop by enough to consider increasing my number of new daily cards. At this point, I want to get to the end of the decks I already have so I can start learning French via flashcards as well. Not to mention, I want to keep the whole experience challenging. With Japanese, I'm far enough along that when I'm not doing something more immersive, I really need to up my game.
I've largely started abandoning all buttons except Again and Great in Anki.
This flies in the face of most recommendations (which recommend avoiding the Great button). But, I think it actually makes perfect sense considering all things like my objectives, number of reviews and so forth.
As I said in the past... unless you're studying for a specific test and need accuracy on a smaller scale, faster, then you actually want less than 100% accuracy. If you're getting perfect scores, that means you're studying at an EXACTLY perfect rate (highly unlikely), or that you're babying yourself (MUCH more likely).
Furthermore, when you're getting perfect scores, it is impossible to measure improvements in your methodology and even harder to convince yourself to seek improvements.
My review load isn't massive. But it is also not small. On a typical day, I have between 150-210 reviews at the moment. Usually somewhere in the middle of that. And it takes me 35-45 minutes. I typically get an accuracy of 95%. Not perfect. But, close.
For me, this is too close to perfect for my liking. I'm still complacent with the results. So, I was thinking about how to "fix" the situation. And the solution I came up with is somewhat counter intuitive.
By preferring buttons which push my cards off for shorter periods of time, I'm maintaining a consistent accuracy and review load, primarily by balancing out babying myself, with the more frequent exposure. In short, there is a balance, and it comes without any cost to me. It is almost a self-maintaining balance. The cards are more frequent, but that also makes them easier to remember, which decreases my odds of hitting "again".
To break this, by preferring Again and the maximum interval, I'm pretty much always punishing myself on the cards I haven't mastered, while properly kicking out the ones I have mastered. I'm a bit more flexible with newer cards (21 days or less). But, once they are out past that, if I get them right, they get kicked far down the road, and if I fail anything on the card, it is back to basics.
If I REALLY do know my cards. Then the review load will drop. If I've been babying myself though, the load will increase and the only solution will be to find better ways to commit the cards I'm failing to memory. And, as I improve my methods for remembering, the load will drop again and faster than it would if I stuck with shorter intervals.
Basically, I inverted the recommendations for the sake of putting the onus for improving on me.
The suspicion here is that at the moment I'm being bombarded by a mix of cards I don't know well enough and ones I know too well. And that isn't a problem in and of itself. But, the ratio of cards I know too well, is too high. And of course that the current rates of change are too forgiving and not putting enough pressure on me.
If I'm right, then after a few weeks my daily review load might drop by enough to consider increasing my number of new daily cards. At this point, I want to get to the end of the decks I already have so I can start learning French via flashcards as well. Not to mention, I want to keep the whole experience challenging. With Japanese, I'm far enough along that when I'm not doing something more immersive, I really need to up my game.
Comments
Post a Comment