Hello,

I just noticed that I keep forgetting some of the earlier characters I learned because the spaced repetition algorithm seems to be "too much spaced"

So, is there a way to go randomly through "old" already learned vocab lists with entirly ignoring spaced repetition , so I can test the characters on a "neutral" way,...and if I dont know them, I click "forgot" and spaced repetition gets an minus (so it would show up more often again) ?

Basic Idea in short:

1. go through already learned vocab lists in random mode (ignoring spaced repetition)
2. adding the NEGATIVE results of "random mode" to spaced repetition

Is that already somehow possible ?

Have you tried studying only a specific list or a section of a list? (There is such a bottom at least in the iOS version). I actually don't know if that would work, but it might be an easy and quick try.

I am studying on PC ,... I also tried single lists, this is possible, but lists are big and "in spaced rep. order" , so its not really working

But thx anyways ;)

@Hemoride: You could try pasting some words from the list in the scratchpad with scheduling on (via the link at the bottom of www.skritter.com/study), and then go through and mark the ones you don't remember well as wrong. That should do the trick!

FYI it is also possible to study just one section of a list from the PC on the website. (In case your big list is divided into smaller sections.)

I empathize with your problem, but if your number of words is already that large the sampling it in a random way may not be satisfactory for you either.

I have never used the "remix" function before, but that's probably another way to excerpt a smaller subset from a huge list.

The thing about SRS is that there seems to be an implicit goal of "never forget any word that you have ever studied." That was a workable goal for me when my vocabulary was a few hundred words and my short term goal was a thousand. Beyond that the time commitment needed to keep retention high seems too much, especially as incremental words are more likely to be "low frequency." So nowadays I am trying to do more purposeful "intelligent cramming." I try to only study lists of words that I expect to encounter in written form in the very near future. It is to early to say whether this is costing me too much in long term retention on the words that I am neglecting, but I hope that over time the "important" words will keep bubbling to the surface if I do it this way, while less important words will be appropriately de-emphasized.

@podster

> The thing about SRS is that there seems to be an implicit goal
> of "never forget any word that you have ever studied."

I have also run into this problem.

> So nowadays I am trying to do more purposeful "intelligent cramming."

I find myself doing the same.

> I hope that over time the "important" words will keep bubbling
> to the surface if I do it this way, while less important words
> will be appropriately de-emphasized.

That is my hope too, but I suspect some changes to Skritter's algorithms would have to be done for that to really work.

It's possible that Skritter would do the opposite: rather than bubble important words to the surface, it would regularly find unimportant words that you have failed at many times and say: "Hey, you haven't reviewed this one in a long time!".

Maybe we need a "not important" button?

This article is pretty good on certain limits of SRS:
http://www.hackingchinese.com/is-your-flashcard-deck-too-big-for-your-own-good/

@ Kai Carver
Thanks for your reply. I noticed recently on my iPhone the "star emphasis" slider in the settings. Maybe this will help, but I haven't tried it until recently so haven't noticed much difference, probably because I'm just doing single lists. I'm not sure how it works (I suppose I could read the directions) but I assume that it is supposed to repeat user-selected "starred" words more frequently than would otherwise happen under the unmodified algorithm.