|
 |
How much time studying vocabulary? | ||
Tags: Time to learn | Vocabulary learning | ||
Share with:
| ||
 Language Learning Forum : Learning Techniques, Methods & Strategies |  | |
| daegga Tetraglot Senior Member Austria lang-8.com/553301 Joined 4310 days ago 1076 posts - 1792 votes  Speaks: German*, EnglishC2, Swedish, Norwegian Studies: Danish, French, Finnish, Icelandic | Message 41 of 350 29 April 2015 at 12:58pm | IP Logged | ||||||||
4 seconds seem ok for an easy deck, not sure about writing with Anki, I don't do it. The other part is that your calculations assume you never fail/reset a card. The usual Anki user will have an error rate of 10 - 20 % per day, it's per design. This will add up if you let Anki reset the card as in the default preferences. So in these (few I guess) cases you do fail a card, what do you do? 1 person has voted this message useful | |||||||||
| rdearman Senior Member United Kingdom rdearman.orgRegistered users can see my Skype Name Joined 5025 days ago 881 posts - 1812 votes Speaks: English* Studies: Italian, French, Mandarin | Message 42 of 350 29 April 2015 at 1:00pm | IP Logged | ||||||||
That's me fecked then. 2 persons have voted this message useful | |||||||||
| smallwhite Pentaglot Senior Member Australia Joined 5097 days ago 537 posts - 1045 votes Speaks: Cantonese*, English, Mandarin, French, Spanish | Message 43 of 350 29 April 2015 at 2:17pm | IP Logged | ||||||||
This article lead me to believe that easy decks are good (and that learning and memorising are different things): Effective learning: Twenty rules of formulating knowledge Have you seen my posts about how instead of forcing in new words, I pick 10 easiest words to SRS out of a list of 2000-4000 words? That makes my cards easy for me and I have a very low failure rate. When I do fail a card, I stare at it a bit, and then reschedule it using the current interval. I don't reset it because that doesn't make sense. Let's say I failed the current interval of 32 days. That means I did get past 16 days last time. So why should I start with 1 day? I think it makes better sense to re-attempt 32 days or be prudent and go back one step to 16 days. I don't think Memrise resets either. Sometimes I give up and put the card back to the 2000-4000 word pile. That's an easy decision to make because I'm not deleting it; it's still there. I'm looking at a backup and I think my error rate there was 3.34%. I remember having an error rate of about 2.5% in Anki for forgot-what-language, so let's just take the more believable 3.34%. I don't know the maths to incorporate that, but I do know that if I started with 8276 words, after an error rate of 3.34%, I'd end up with 8000 (correct) words. Then, 4 seconds x ( 6 initial reps + 9 review reps ) x 8276 cards / 365 days = 23 minutes per day answering cards 0.76 minutes more. Insignificant. --- In fact, "SRSing 8000 words in 1 year at 22 mins/day" seemed outragious(?) for many, but what if I used a more Joe-friendly "SRSing 8000 words in 2 years"? Anki doesn't cap intervals but Memrise does, at 180 days. ie. their intervals are: 4 hours, ..., 90 days, 180 days, 180 days, 180 days, 180 days, 180 days, ... So, over 2 years, you do 17 reps ( 6 initially + 9 in first year + 2 in second year). 4 seconds x ( 17 reps ) x 8000 cards / 730 days = 13 minutes per day answering cards Just 13 minutes. It's actually much less. Using 15%, average error rate you said Ankiers usually have, we learn 9412 words to start with, and the maths become: 4 seconds x ( 17 reps ) x 9412 cards / 730 days = 15 minutes per day answering cards Now that's Joe's figure, not mine. Edited by smallwhite on 29 April 2015 at 2:19pm 2 persons have voted this message useful | |||||||||
| Jeffers Senior Member United Kingdom Joined 4698 days ago 2151 posts - 3960 votes Speaks: English* Studies: Hindi, Ancient Greek, French, Sanskrit, German | Message 44 of 350 29 April 2015 at 2:19pm | IP Logged | ||||||||
I agree, and it raises another interesting point about how we use flashcards. What do we do when we look at them? When I look at an Anki card I pronounce the word carefully, usually in my head. That alone takes me 2-3 seconds in most cases. I look carefully at the word's spelling, I recall cases where I've recently come across it, I try an example sentence or two in my head, picture the object referred to, or just think about how the word is used. My response times on Anki therefore range from a minimum of 4 seconds to anything up to 30 seconds per card. If I just can't remember the word, or get it muddled up, I usually fail it which puts it back to being a new word. Edited by Jeffers on 29 April 2015 at 2:21pm 3 persons have voted this message useful | |||||||||
| chaotic_thought Diglot Senior Member United States Joined 3331 days ago 129 posts - 274 votes Speaks: English*, German Studies: Dutch, French | Message 45 of 350 29 April 2015 at 2:38pm | IP Logged | ||||||||
|
I'm trying to do more of my vocabulary learning nowadays with Listening-Reading, sometimes called "L-R". Learning takes place when you actually read and listen to the text. The purpose of reviewing the vocabulary is 100% to review not to "learn". Since we're talking about efficiency, I thought I would take a few random entries from my L-R sessions and see how they look in their most "efficient" form:
penning / badge hufter / bastard forsch / powerful gebouwd / built goedheid / an act of mercy This list is efficient, but glancing at the list, the problem I have is that it doesn't bring to my mind any of the situations where those words were used in the texts. Looking at this list is exactly the experience of looking at some ink splatterings on a page which have been digitally rendered on a computer screen. I've tried "cramming" such lists e.g. using Anki before and basically got the following results: (1) I always end up forgetting the same ones over and over, (2) reviewing the words is no fun at all, and finally (3) no real sense of accomplisment or progress. So instead I'm going to try the following way. I notice that if I include the entire phrase from the L-R session, my impression from looking at the list changes drastically. I can not only recall the situations where the words were used months after having read the book, but I can even "hear" and "see" each character speaking and acting out as I imagine they would have if they were real: Je raakt je ____ kwijt. / You’ll lose your badge. Wat ben jij een ____. Weet je dat? / You’re a bastard. You know that? Ned Land was ____ ____. / Ned Land was powerfully built. Volstrekt niet, mijnheer, dat is ____. / No, sir, it’s an act of mercy! For reviewing such a list, Anki will not work well because it will destroy the chronological order of the list. Anki will also insist that you perfectly remember every single item, even if some items are obviously not as important as others (e.g. "hufter" from the above list). Instead of Anki or some tool like that, simply format this list in a convenient way and review it using "week doubling" (e.g. first review date is on week 1, next one is on week 2, then on week 4, week 8, ...). After each review session, make sure your recollection rate is no worse than it was during the previous review session. If your recollection worsens, schedule a catch-up review session for the next week so that your recollection rate catches back up. After "doubling" a few times, it will quickly become obvious to you when you have "committed" the items to your memory. For example, if you can answer the list with 98% recollection even after 32 weeks of not having looked it, is there any doubt left in your mind whether you can remember those items? If you print the list with the unknown items formatted in a particular color, then you can use a colored transparency to "conceal" those words to test yourself during your review sessions. To determine your recollection rate, count up how many lines you could remember out of the total (e.g. 80 out of 100). If you are used to working with HTML, you could alternatively arrange the list on an HTML document so that the unknown words reveal themselves only after you click on the blanks. In this format you can also set up a simple form so that you can count up how many lines you remembered correctly. Each time you remember one of the entries, check the associated checkbox and then at the bottom of the page print the total count (e.g. 80 out of 100 items checked). Edited by chaotic_thought on 29 April 2015 at 2:40pm 1 person has voted this message useful | |||||||||
| smallwhite Pentaglot Senior Member Australia Joined 5097 days ago 537 posts - 1045 votes Speaks: Cantonese*, English, Mandarin, French, Spanish | Message 46 of 350 29 April 2015 at 2:50pm | IP Logged | ||||||||
I know that's the most intuitive thing to do. But is it productive? Let's think about it. Will doing that IMPROVE MY LANGUAGE LEVEL, taking me from, say, B1 to B2? No, because I don't learn anything new from it. Will doing that IMPROVE MY MEMORY OF THE WORD? I believe so. But it's taking ~3 times more time, thus 66 minutes per day instead of 22 minutes. I'd rather do 3 more reps of the word, or use the extra 44 minutes per day to read, or learn 3 times more words, or learn 3 more languages, or sleep. Is it NECESSARY for you? Did you compare doing it and not doing it? What are your goals and does doing that let you meet your goals faster? Edited by smallwhite on 29 April 2015 at 3:03pm 2 persons have voted this message useful | |||||||||
| smallwhite Pentaglot Senior Member Australia Joined 5097 days ago 537 posts - 1045 votes Speaks: Cantonese*, English, Mandarin, French, Spanish | Message 47 of 350 29 April 2015 at 3:13pm | IP Logged | ||||||||
I don't know why the difference, but my SRS cards look like that yet most of them do remind me of the storyline/context. Prepositions and adverbs maybe not always, but nouns and verbs certainly do. I wonder what you prefer: * that single word cards don't remind you of their original context, and you do sentence cards instead or * that single word cards do remind you of their context, and you do single word cards ? Which would you prefer in an ideal world? Edited by smallwhite on 29 April 2015 at 3:14pm 2 persons have voted this message useful | |||||||||
| tarvos Super Polyglot Winner TAC 2012 Senior Member China likeapolyglot.wordpr Joined 4496 days ago 5310 posts - 9399 votes Speaks: Dutch*, English, Swedish, French, Russian, German, Italian, Norwegian, Mandarin, Romanian, Afrikaans Studies: Greek, Modern Hebrew, Spanish, Portuguese, Czech, Korean, Esperanto, Finnish | Message 48 of 350 29 April 2015 at 5:13pm | IP Logged | ||||||||
|
Fors, not forsch. And "fors" doesn't mean powerful usually, it usually means large.
2 persons have voted this message useful |
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
