So I noticed something today about my kanji reviewing. It seems that when I review expired cards by clicking the "review all expired cards" button, that my retention rate (or percentage) is much, much lower than if I click on the orange cards stack by stack.

I'm talking 95% average when clicking on the orange stacks separately, and 70% when clicking on "review all expired cards."

Is there someway you're suppose to do it? As in, you should be mixing them all together, or is it OK to study stack by stack and pass them. I guess it could kinda be viewed as cheating because, once you see what it was from, the primitives for that lesson pop into your head and you can just use those to get them all right.

But I hate having a 70-75% retention rate.

I would recommend doing the method that gives you the lower retention rate. This would seem to be more close to your "actual" retention rate. It feels good to pass the cards, but what's the point if you don't really know them well?

I have noticed the same effect. My guess is that cards in each stack are more similar because of identical primitives, and because they were learned at the same time your brain recalls them more easily together. It's like reviewing a list of vocabulary on a piece of paper where the words are always in the same order. It's much easier. Taking them out of context and presenting them randomly is more difficult, but a more accurate test of retension since in real life words will come at you more randomly. I suggest clicking "review all" and just biting the bullet on retension rate until you can bring it up with more practice.

Ahh, oh well... Looks like I'll just be using that "review all" button. Maybe if I ever get into a slump and need a motivation boost, then I'll go ahead and review stack by stack .

I do "review all" as well, but there is one good point about doing stack by stack that might be worth considering;

if you do stack by stack, you will know which stack/time point it is that you have the hardest time with. For example, if your stories are fine for the first couple of reviews, but don't last til the third and fourth, this is a different sort of problem from just not remembering cards that you have only reviewed once. Then you can fix your stories accordingly.

Hmm..
All the cards get shuffled together anyway when they reach the latter compartments so is reviewing stack by stack really counter productive or is it actually better since it's easier to remember them that way. It might even be better for enforcing the mnemonic because it still comes "keyword to kanji" but when you have to check it "kanji to kanji", something is lost in the process.

I'm thinking, it is a bit softer ride, so it might be better for your motivation and you still get to the same point when all your cards are in the latter compartments. There you'll have no shortcuts and the characters that really need some work will drop out at that point in the latest.

Edit: wrong topic, sorry.

I guess it kinda depends on when you'd prefer to have your problems, now or later. For me, failing a card is just the opportunity for additional review. I don't like the idea of reviewing individual stacks because I'd rather just figure out what kanji I have problems with ASAP so I can fix the issues and adjust my stories right away before an ineffective story burns its way into my brain and gets stuck there.

I'd argue that it's whether you'd prefer to have your problems now or never

It's highly pace-dependent, of course, but since I switched to stack-by-stack reviewing I haven't noticed my retention rate in the top stack going down at all, while the higher pass rate means that I only get somewhat butchered by massive timeouts of cards in stack two.

~J

Well, definitely a bunch of good points here. I've just been sticking with the "review all" button. My rates are climbing back into the 80%'s (probaly because my card input has slowed a bit recently), and it seems to be working ok.

However, that whole thing about review stack by stack to identify problem stacks is definitely worth considering, I think. But I'm going to keep trucking this way and see how it goes.