Hello, I need help to figure out what’s the best choice.
I’m using Anki to study physics, and I have the manual of the class in pdf.
To create cards, I could either copy the text / the formulae from the manuel, or take a screenshot and past it in Anki.
I can gain a lot of time by making screenshots. But, I know images have a bigger size, and need to be synchronized. Also, typing and creating the card by myself can help memorization.
So, I don’t know which option I should prefer.
I don’t think that there is evidence that typing in the card contents yourself helps memorization (in the long term). Active with recall should take care of the memorizing. But typing in the questions yourself might help creating better cards.
I first used Anki (for something outside languages) in my Physics masters course, to prepare for the exam in experimental particle physics. We had the lecture slides as a PDF. Sometimes I would just screenshot an entire slide, put it on the answer side and create a question like “describe and sketch out Wu experiment”. That was not good. If I missed a single part of the description, or made just a small mistake, should I mark it again, hard, or good?
Later I learned about making shorter, more atomic cards, where each should only require the recall of a single atomic piece of information. Check out Supermemo’s 20 rules for formulating knowledge and Andy Matuschak’s How to Write Good Prompts. From that I learned that creating good cards often requires rephrasing the source material. Sometimes a single sentence in a text becomes several cards, sometimes you need to create cards for what’s implied between the lines etc.
Generally, I put more thought into the question “design” when I type the questions manually. It also gives me time to think whether I understand the question. However, I agree that you can safe a lot of time by using screenshots of formulas instead of typing in the LaTeX yourself. For aesthetic reasons I prefer typing in formulas by myself, modern versions of Anki have really good MathJax equation support and with practice you can get fast at it. But I don’t think that typing in equations helps much with long-term memory.
One very practical usage of images are image occlusion cards. In Physics I don’t think you can use them a lot, but I could imagine some cases where they could be used, e.g. for labeling Physics experiments or other diagrams. With image occlusion you can get many atomic cards from a single source image.
If you are worried about image size, you can reduce that by converting them to the webp format, for which there is the Paste Images as WebP addon, which might help a bit.
Thank you. I will try both methods to see if its really worth it. As you pointed out, the longest part isn’t the creation of cards but rather the study.
As you pointed out, the longest part isn’t the creation of cards but rather the study.
I don’t think this is necessarily true. Andy Matuschak wrote in his guide:
An easy prompt will consume 10–30 seconds across the entire first year of practice, and much less in each subsequent year.
Tbh, I think that he is quite optimistic here, but it should not be several factors more. It’s quite realistic if you take few seconds to answer per card and answer correct most of the time. But this is under the condition that you create good cards which just ask for a single piece of knowledge. Don’t make cards where you screenshot an entire paragraph and then create a question which asks you to basically say everything in that paragraph, because that often will be multiple pieces of information.
I used screenshots often for formulas, which I think if fine. However, for long formulas (like the Bethe–Weizsäcker mass formula) you might also consider splitting them up into multiple cards. E.g. for the gauss function one card for the exponent and one for the factor in front of the exponential. You can also use cloze-deletion for splitting equations into multiple parts, however this only works with text. Or you use image-occlusion on the screenshot of an equation (I never tried that).
My replies are a bit long and a bit vague because from your question it had not been clear how you use images and screenshots.
What I meant regarding the creation of the cards was, that I don’t think the act of typing helps much with long-term memory. It might help with short term memory. But repeating the card with Anki will make sure it gets to long term memory even if you never type it in yourself. But that does not mean that the Anki recalls will take more accumulated time. You can repeatedly write down a new formula for an entire hour and still not remember it next year if you don’t do any revisions in the meantime.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.