Why not have an image that is partially masked out (in a different place each time) and have the user fill in the missing pieces? But in the digital realm the possibilities for modeling sequence completion and contextual infill are much broader. One, extremely limited, way you can model that is with the front and back of physical pieces of cardboard. Given a context what comes next, what is missing? That is only suggested by the medium of physical flashcards and not the learning task itself. There’s no reason that we need to have a front / back pairing for information. (4) The process is constrained by the flashcard metaphor (3.3) This leads to significant frustration and demotivation (3.2) Thus I am constantly failing to recall and having to relearn information (3.1) It is not easy to tune these curves in the apps. This might not be the case if (2) were improved, but it is for isolated factoids. Going from 1 to 3 months? Insane! I need a much much more dense series of practices all along the curve to hope to recall information. The default spacing in repetition apps I’ve tried doesn’t work for me at all. (3) I forget faster than the built in curves predict. If the method of creating cards were primarily through other modalities and secondarily through text input it might force designers to make those primary modalities easy to use. (2.3) Thus knowledge learned through Anki is fragile and shallow. (2.2) Thus important context is not captured or practiced. (2.1) Thus the dense connections provided by imagery, audio, video and the act of physically writing things out are absent.Įvery act of retrieval that is limited to declarative knowledge stored as language is going to be more prone to failure than one with multiple encodings across sensory and motor modalities. The best practices suggested for creating cards is that they are short pieces of text. ![]() This means you have all the limitations of keyboards. The primary methods of input to create Anki cards is typing. (2) Easy input is constrained to short snippets of text If a component of this process presents a huge wall of effort I am going to relogate it to the pile of disfunctional tools. The entire point of spaced-repetition for me is to minimize effort in relearning things I’ve forgotten. If you don’t set that cap you can cover all the cards but it means that a given session has an unbounded amount of effort required to complete it. ![]() While Anki can limit the number of cards you practice during a given review session all this means is that there are a ton of cards you haven’t reviewed and are thus not learning or are forgetting. (1.1) Because there is no input rate limiting your backlog can grow at an unbounded rate. I’ve created dozens of new cards in a sitting without breaking a sweat, but this is ultimately self-defeating. This is easy to do when you’re diving into a new subject. I may work on creating a tool that doesn’t suffer from these defects later, but in the meantime I hope others can suggest alternate programs that do not suffer from these drawbacks.Īnki will let you add as many cards as you want to a deck at a time without warning. In this post I want to capture why spaced-repetition, and specifically Anki, doesn’t work for me. I almost always stop using it after a couple of months at most. I love the idea of spaced-repetition and I have attempted to use it several times but it has never stuck.
0 Comments
Leave a Reply. |