All Collections
Card Design
Writing Atomic Flashcards
Writing Atomic Flashcards

For more efficient, enjoyable study sessions, make your flashcards as clean and small as possible.

Soren Bjornstad avatar
Written by Soren Bjornstad
Updated over a week ago

When doing any kind of active recall study, but particularly when studying flashcards with spaced repetition, the most important rule for writing flashcards is this: the smaller your flashcards, the more efficient and effective your study will be. To be more precise, each flashcard should be atomic.

What is atomicity?

What do we mean by atomic? In short, an atomic flashcard can't be broken down into multiple parts which are understandable and useful on their own – it's a single unit that tests your knowledge of exactly one piece of information.

For instance, suppose we wanted to learn about the Python programming language. The most obvious approach would be to create a flashcard that asked you to describe the language:

This is not an atomic card. It has at least seven distinct facts on it – maybe more, depending on how you count! Let's rewrite it instead as a series of atomic cards:

This rewrite uses the Concept-Descriptor Framework, but if you prefer, you can just as easily write each of the indented descriptors (the parts in italics) as questions, like “What organization maintains Python? → The Python Foundation.”

Why does atomicity matter?

So sure, you say, there are seven facts in one flashcard in the first example, but why is that bad? In short, this card will unavoidably produce inconsistent recall and be frustrating to practice.

First of all, consider: if you come up with fewer than seven of those facts, how are you going to rate yourself? If you said that Python was “a high-level, multi-paradigm language written by...”, and just left out the “interpreted” part, do you know what Python is? What about if you forgot the name of the creator? Or you left out one of the common uses? You'll end up using your limited willpower and study time figuring out what counts as knowing what Python is, rather than actually learning the facts – or maybe you'll just say you got it and move on, even when there are facts you originally wanted to know that you forgot.

If you're like most people, you'll end up picking up on a random subset of the information on the card – whatever's easiest to remember – then always respond with that when you get the card and figure that seems good enough. You'll then completely forget the rest, even though you have that information as a flashcard in RemNote! Then when it comes time for an exam, you'll discover you randomly don't know some of the information you were trying to learn. This leaves you with no control over what you actually learn ­– which is the entire point of studying with RemNote.

You'll also never have the satisfaction of feeling “yes, I know that!”, unless perhaps you exactly memorize the entire paragraph word-for-word and recite it when you see the card. That's a big problem, because getting things right while practicing flashcards is a primary source of motivation; you'll get tired and burned out much faster if you don't feel this.

Lastly, because RemNote cannot keep track of how well you know each of the seven pieces of information individually, even if you somehow do actually remember all of the items every time, you'll need to spend extra time studying to get good results. That's because when each fact is scheduled separately on a separate flashcard, you can review the difficult ones frequently and the easy ones only occasionally. When all the facts are slammed together onto one flashcard, you'll either have to study that flashcard as often as the most difficult item on it, or repeatedly forget the most difficult items.

Portions of this page were written with reference to Control-Alt-Backspace's series on memory (also written by the author of this article).

Did this answer your question?