SRS question

I’d like to use the built in SRS review system instead of Anki for convenience sake but I don’t understand how the built in system works. Some clarification regarding the following would be much appreciated:

How does LingQ determine the intervals? What’s the formula?

Does LingQ introduce new words for review for me like Anki does or do I just go in and start a review session? Is there a setting where I can adjust the amount of new words per day etc?

If I review multiple times per day will this break or confuse the SRS formula so it thinks I know a word really well and not test me again for a long time (if at all)?

It’s a hassle for sure exporting words to Anki but that’s what I’m doing since I know how Anki works I feel much more in control and I have faith that it works.

Speaking of having control over what I study. I am probably using LingQ a lot differently than most users and I really like this method so I thought I’d share. I have it set to mark all unknown words known as I page. I do not create LingQs for every unknown word rather only the words I want to study and then I continue up until the daily limit. I think it’s set at 13. I’d like it if I could manually adjust this but 13 is a pretty good number I think anyway so that works out. So then in theory I now have 13 new words to learn every day added to my SRS. My “known” word count obviously gets super inflated but since LingQ doesn’t count words as they’re listed in the dictionary but rather every conjugation is its own word so to speak I don’t find the number too useful anyway.

2 Likes

I agree with you!!! I would like to see a significant improvement in the SRS mechanism. ANKI’s concept is excellent and I think it’s so importants that LingQ will learn from ANKI and implement functions from ANKI in the SRS.

2 Likes

My big problem is that there is apparently no way to make different decks. It seems like it is always drawing words from your total list on LingQs. And very quickly the total list gets waaaaay to long to manage on an SRS system. I would love if I could make a deck of a couple hundred LingQs that I want to review and focus on those with SRS.
Currently I am just exporting my important LingQs to Memrise for review, but it would be nice if I didn’t need to do that.

1 Like

Regarding the SRS, every day words and phrases are added to your Due for Review list. You can find this list in the Vocabulary section. LingQs are selected for review based on our SRS (Spaced Repetition System) schedule. Status 1 - 1 day, Status 2 - 3 days, Status 3 - 1 week, Status 4 - 2 weeks, then 1 month, then 3 months.
Daily lists are generated from the list of terms due for review. You can set the number of terms that get sent each day up to 200. Of course, if you go to the Vocabulary section, you can review all your terms until there are none left for review.

In fact, we don’t recommend this type of review for all your LingQs. You should be creating more LingQs than you can possibly review this way. It’s far more efficient to learn this vocabulary by seeing it in multiple contexts over time. The important words will reoccur more frequently and will be learned soonest, and the other words will be learned over time. In effect, you review naturally as you move on to new lessons.
Reviewing this way, in context, is actually far more efficient and will allow you to learn the tens of thousands of words you need much more quickly and in a way that you will actually retain them.

Having said that, the review tools are there and can help you get started, or for a change of pace sometimes to do something different while still learning. Or, to help you focus on specific vocabulary lists which you can create by tagging LingQs and then filtering for these tags in the Vocabulary section. Do what you enjoy but heavy memorization is not really necessary.

2 Likes

I don’t use the SRS much, but: wouldn’t using tags be the best way to do what you want – ie focus on subsets of LingQs?

Possibly. As far as I can tell there is no way to add tags to multiple LingQs at a time, it has to be done one by one. At that point it’s just a lot nicer to export them to some other system that gives more flexibility.

Ah, I see. That is a pain! I just checked and can’t find a way to tag multiple items either. If I were more invested in using flashcards, I would probably export to Anki.

But hey, before you put more time into this, consider watching this video (made by a LingQ user) about why flashcards are not the best way to study:

I agree that flashcards should not be a main learning activity. I definitely do much more reading, listening, and watching than anything else.
However, after learning the first several thousand most common words, the incidence of the other words falls off very quickly. In my experience supplementing reading and listening with SRS at this upper intermediate stage can be very useful. It definitely helped me break out of a plateau in my Spanish.

1 Like

This video is deeply wrong from start to finish.

“when the vocabulary starts to accumulating SRS review will take up pretty much all of the study time.”

My deck contains 10575 cards and the daily review takes 36,5 min on average. I use my phone for Anki and spend my dead time to do my reviews. (not at once)

I wish people would focus on promoting their own method, if they want to, and would stop bashing other people’s methods. It’s not a coincidence that Anki is recommended by a lot of people.

I’m not sure if the video is “deeply wrong from start to finish”. But perhaps it’s more opinionated than it needs to be, and does not leave much room for a strategy that incorporates any amount of flashcards.

I used to be all-in on flashcards (Anki, +/- Fluent Forever), for a number of years, and did not see much improvement in my language skills. This video (and various of Steve’s) helped get me started with lots of reading/listening, which I’ve found more enjoyable and productive.

You seem to have found an approach to language learning that incorporates reading/listening and also flashcards. I’d be interested to hear how you balance these activities, if you care to share. Particularly I’m curious about how you decide what to add to Anki, and in what form – ie, do you use simple term/definition cards? or do you get fancier w/ sentences, clozes, images?

My study method is reading/listening just like yours. I use anki as a tool to lock in words faster.

Regarding my reading, I have to mention that I have a high tolerance for boring stuff, so it’s easy for me to find content for any level. But I refuse to read anything more than once, and I hate ambiguity. If I see a new word I want to learn it immidiatelly. So my method is “intensive reading” instead of “extensive reading.”

But these are my own experiences. I don’t make a youtube video about why extensive reading is not an effective method of language learning. Just because something doesn’t work for me doesn’t mean it’s not effective.

The quality of the cards is not important. Anki works because of the algorythm behind it. Because of the SRS and Active recall approach. I think the biggest misconception is that learning words without context is a bad thing. People say it’s easier to learn a word with context. This is true because the context triggers the meaning of the word. But the point is not to learn words easier. The point is to be able to recognise the words while reading new content. You see the context while reading anyway.

The important thing is to be able to learn many words in a short time with a reasonable retention rate (80-90%). So I use basic cards, without context. If I see a new word and know I won’t recognize it next time, I’ll make a new card out of it.

I usually read for an hour every day. I copy / paste difficult words into an Excel document while reading. When I finish reading, I import the excel into Anki.

For me, doing anki is not an active study time. I’m just washing my mind with the words and I know it will stick. This is an opportunity to come across words often. And like I said, I use my dead time for Anki.

A few points: No one should use the default settings. The value is in the algorithm. The default settings cannot take advantage of the true value. No one should press the hard or easy button. These two buttons mess up the algorithm. The retention rate should be between 80-90%. If the retention rate is below 80%, you forget too much. If it’s over 90%, you’re not learning enough new words.

The amount of new words a day is up to you. In my experience, 50-100 words are manageable. In fact, 50 new words were pretty comfortable. My average review time is 5 sec/card. And creating the cards is fast because of the “copy/paste method to excel and import later.”

1 Like

Sorry for the slow reply, but I wanted to thank you for sharing your thoughts about this. It is especially helpful to hear the mechanics of your system (copy/paste into Excel), and your advice that “The quality of the cards is not important.”

In the past I put too much effort into making “high-quality” cards, and now feel much of that was wasted effort. (The Fluent Forever theory that browsing a google image search in your target language will help you remember that language just seems silly in retrospect.)

But you are right, it’s the algorithm that is more important, and the ability to use downtime to flip through selected vocab and learn it through spaced repetition: more quickly and efficiently than waiting to encounter these words through reading.

I still think it’s best to prioritize reading and listening (which it seems is your primary activity too). But I’m planning to start incorporating flashcards, with an approach like the one you describe. Perhaps I will report back about how it goes.

Actually I have one more question for you: do you use comprehension flashcards only (with the target language on the front)? Or do you also use production flashcards (with the target language on the back)?

I only use comprehension cards.

Production cards are harder. They require more time. If I had more time to invest in anki, I would learn more comprehension cards. My goal is to condition myself to be able to recognise words quickly. That way I can read more fluidly.

If I want to work on my active vocabulary, I usually just write a lot. So basically, I activate my passive vocabulary by using the words.

Anki can help to activate passive vocabulary, but it’s time consuming. The number one rule is to do all the review cards every single day. (That’s why Anki isn’t for everyone.) Anki doesn’t work without consistency.

So it’s important to manage the load. Maybe if I were a full-time language learner, I would also use production cards.

1 Like

Hi kraemder. Sorry if I ventured off topic in my comments below about the overall utility of flashcards. Here I want to add one question to those you raised in your original post, about how LingQ’s SRS system works.

Background: When I click the flashcard button within a lesson, I see options to “Review Due” and “Review Lesson”. The number of flashcards available for “Review Lesson” is always greater than that for “Review Due”. This would make sense if I were regularly reviewing flashcards and/or giving my LingQs a range of status from 1-4.

However: I almost never actively review flashcards on LingQ, and the majority of my LingQs either have status “1” or status “known”. Per Zoran’s explanation, this would mean that nearly all of my LingQs should be “due” every day. And so the number of flashcards available for “Review Due” and “Review Lesson” should be pretty much the same.

So my question is: Why is the number for “Review Due” and “Review Lesson” always quite different, if I never review flashcards and only uses status “1” or “known”?

A related question: What counts as a “review” on LingQ? Do I have to engage in a flashcard review session? Or change the status of a term? Or are there other activities that count as “reviews”, and hence alter the due date for these words and explain the discrepancy noted above?

1 Like