Allow the button HARD to advance to the next step in learning state (slower than the button GOOD does)

I have been tying to train my recognizing of musical intervals by ear and use Anki to schedule and perform some if my repetitions. Recognizing intervals is quite difficult for me. Often I grade my answers as “HARD”. I configured many “steps” for new and lapsed cards in that deck. However, if I rate my answer as “HARD” when a card in “learning” state, the card could get stuck at the same step indefinitely, because I need many months of repetitions until I feel like grading my answer as GOOD.
As a workaround, I do not use “HARD” button when the card is in “learning” state. When I see, that the card is in “learning” state, I use GOOD instead of HARD.

I suggest adding two checkboxes to Deck’s configuration in New Cards and Lapsed sections:

  • Checkbox 1: GOOD is required to move to the next learning step in “learning” state.
    • if checked it works as the current behavior, would be the default option: GOOD advances one step ahead, HARD retains the card at its current step.
    • if unchecked, GOOD still advances to the next step; HARD advances by a “half-step”. Selecting this button two times for the same card would be equivalent to selecting the button GOOD one time for that card and will advance the card one step ahead.
  • Checkbox 2: GOOD is required to finish “learning” phase at the final step.
    • if checked it works as the current behavior (would be the default option): GOOD finishes the “learning” phase and set the interval and due date for the next repetition. HARD retains the card at the final step of learning phase.
    • if unchecked, both GOOD and HARD finish “learning” or “re-learning” and set the next due date for review.

Recognizing sounds is just one example of inherently difficult material. I can think of more examples:

  • Training to speak a foreign language with a proper accent or training to speak with a specific accent in your native language
  • Anki could be used to build a skill, that requires muscle memorization of fine or very fine motor skills, like speed blind typing on keyboard, singing, dancing, playing musical instrument, performing a brain surgery in simulation (requires additional software and hardware), speedcubing etc.
  • Any material could be difficult for a person with memory-related health problems, see this study.
  • Training for a mental calculation competition

The idea is that a card doesn’t graduate until you can answer it well. If answering a card is still hard for you, why would you want it to become a review card? The learning process is clearing not finished, yet.
On the other hand, if you can answer a card easily, learning is apparently finished and you might just as well skip to the review phase.
I think the current behaviour is quite intuitive.

4 Likes

.

.
I concur with this post in its entirety.
.

.

.
Maybe each card contains too much information. Can you post some images of the fronts of your cards,
.

.
Either (a) too much info on one card, or (b) you need smaller steps.

Can you post your steps for Learning and Lapses.
.

.
Fair enough. But the intervals between steps are also important. Can you post your steps for New Cards and Lapses.

.
So what do you do if you don’t press Hard?

Thanks.
.

Sometimes the learning material could be inherently difficult and could not be simplified.

The front of my card is the sound of a musical interval
Basically it is the sound of two musical notes playing in sequence: the first note and then the second note.

The back of the card is the name of the interval. A name of the interval could look like: “Minor second”.

The steps I configured for new cards are “60 120 180 240 300”.
The steps I configured for lapsed cards are “60 120 180 240”.

The “learn ahead limit” is set to 400.

As a workaround I press “GOOD” instead of “HARD” when I see that the card is in (re)-learning state.

The current behavior works well for normal learning material. But when the learning material is inherently difficult, like recognizing of musical notes or intervals is for me (see my previous reply to [OldGrantonian])(Profile - OldGrantonian - Anki Forums), the user would not feel like pressing “GOOD” for months until a skill of recognizing gradually improves.

Another example of inherently difficult material would be foreign accent training.

1 Like

I suppose you know Ear Master and other similar software, but if not, I strongly suggest you to try them.They don’t use SR, but they offer a lot of flexibility, customization and options for ear training. You might find they work better than anki (or any other SR program) for you.

1 Like

Recognising sounds is not a matter of memorisation. Personally, I think it’s great if you can utilise Anki for it nonetheless, maybe even write an add-on. But I don’t think vanilla Anki can or should strive to solve this problem.

Thank you. I am already using both Anki and Perfect Ear for this. The combination of both works better for me. Once I mastered an exercise in Perfect Ear, I move on to the next one. And I use Anki for repetition of the stuff I already learned in Perfect Ear.

1 Like

.

.
The iPhone has “Tone Deaf Test” and “SingTrue”, et al.
.

Addon is not a solution for me, because I use AnkiDroid.

1 Like

The following link says “Auditory recognition memory is inferior to visual recognition memory”.

Here’s a quote:

“It is clear from these results that auditory recognition memory performance is markedly inferior to visual recognition memory on this task. Note that we do not claim that long-term auditory memory, in general, is impoverished. Clearly, some form of auditory long-term memory allowed our participants to identify the stimuli as tea kettles, dogs, and so forth. Moreover, with practice, people can commit large bodies of auditory material (e.g., music) to memory. The striking aspects of the original picture memory experiments are the speed and ease with which complex visual stimuli seem to slide into long-term memory. Hundreds or thousands of images, seen for a few seconds at a time, are available for subsequent recognition. It is this aspect of memory that seems to be markedly less impressive in audition.”

.
https://www.pnas.org/content/106/14/6008#:~:text=Two%20explanations%20suggest%20themselves.,different%2Fsmaller%20than%20visual%20memory.
.
You say:
.

.
These steps would be far too long for visual items in the Anki Learning phase. So, for auditory items, they would seem like centuries.

According to @giovannihenriksen, you can use fractional steps:
.

.

Recognizing sounds is just one example of inherently difficult material. I can think of more examples:

  • Training to speak a foreign language with a proper accent or training to speak with a specific accent in your native language
  • Anki could be used to build a skill, that requires muscle memorization of fine or very fine motor skills, like speed blind typing on keyboard, singing, dancing, playing musical instrument, performing a brain surgery in simulation (requires additional software and hardware), speedcubing etc.
  • Any material, including visual, could be difficult for a person with memory-related health problems, see this article.
  • Training for a mental calculation competition

Setting the issue with memory-related health problems aside, none of these is about memorisation; they are about practise.
I just leave this here: Law of the instrument - Wikipedia

You can call it practice, or Procedural memorization - Wikipedia, it does not matter. I argue, that Space Repetition is an effective learning method for both. I never said, that you should replace practicing the skill in real life with Anki, it is just an additional tool for more effective learning.

Some references:

Even if you can leverage spaced repetition to some extent for this kind of learning, it’s still inherently different from what Anki has been designed for. Anki should focus on its default use case and not compromise to cover some loosely related areas.

1 Like

If I set so small steps, I spend most of the time pressing AGAIN in the loop of the most difficult intervals (leeches) and do not have time to learn anything else. I do not believe it is an effective way of learning.

It is fine. Fortunately, Anki could be used for inherently difficult material learning as it is. The problem i suggested to fix is just a minor inconvenience.

However, I would like to know, if Anki is “intended” for learning the following type of cards: Minimal pairs. Basically the Front of the card is the sound of an English word and the answer is the word itself. So you have to recognize the word by the sound of it. Sounds easy for a native speaker, but it is sometimes difficult for me, as a Russian speaker to distinguish between “man” and “men”, “end” and “and”, “done” and “don” (in American English). The reason I cannot easily hear the difference between the sounds in those examples is that the different English sounds in each pair “æ” and “e” or “ʌ” and “ɑ” are both similar to only one of the Russian sounds. My point is that sometimes recognizing foreign language sounds could be just as difficult as recognizing musical intervals. So it would be an “inherently difficult material”.

I don’t think that’s a good fit for a flashcard, either. In my experience, you need lots of input to improve your hearing comprehension. If you can distinguish two sound files, it doesn’t necessarily mean you make the distinction based on the differing vowels.
Rating is also hard because you will be right 50% of the time, anyway.
But most importantly, I think, the learning process is completely different from memorisation: If I look at the front and back of a memorisation flashcard, and then get shown the front, I will be able to give the correct answer. The question is only for how long that memory will last.
On the other hand, if I listen to a minimal pair and get told which sound is which, it doesn’t mean at all that I will be able to answer correctly, even if I get to hear the sounds again immediately.
However, once I’m able to distinguish the two vowels, I won’t lose that ability as long as I’m occasionally exposed to the target language, so I don’t need any flashcards.

1 Like

I use Ankidroid special paid version which has pass or fail. I also use text to speech only. I never use eyes to review. Pass or fail even more important for people who only listen to review walking down the street.

The hard and easy buttons are an awful design. First of all if you spend 2 seconds each time to think what to press that’s 2 seconds you could have done in my case 0.3 cards. Many times I caught myself thinking about pressing hard for more than 30 seconds. That’s stupidity.

You either know something or you don’t know it. Don’t do it. I highly advise paid Ankidroid. (No advertising)

1 Like