How did I remember nearly everything I learned in medical school, so that I could score a 270 on the USMLE Step 1 and receive honors on all of my 3rd year shelf exams? By building a strong foundation of knowledge that I never forgot, which I put into Anki flashcards, which to this day I still review. Here I share the method of making cards that can be the difference between toiling away for hours in frustration with little to show for it vs. scoring a 270+ on the USMLE Step 1.
3 key principles (and 4 mistakes to avoid) for Anki success
While I am a strong proponent of using Anki for more than just memorizing flashcards, the way that I recommend most people to start is to treat it as a basic flashcard program. Here, I lay out 3 simple rules for making BASIC flashcards, along with the 4 most common roadblocks that delay people reaching their full potential.
Rule #1: Keep the amount of information on a card to a minimum (ideally 1-3 unrelated facts)
By far the most common mistake that people make when they start out using Anki cards is that their cards contain way too much information. This point is worth repeating, because making cards that are too long hurts you twice: once, because you will have to repeat these cards many more times, which wastes your time as you repeat facts unnecessarily; and twice, because we won’t remember the information as clearly.
Let me give you a real-life example:
On the front, this student had as the question:
Differentiate the infections by the causative agent and describe the rashes in their locations direction of spread. Also describe unique findings.RubellaErythema infectiosumRoseola infantosumHand/foot/mouth dzScarlet fever
And as the answer:
Rubella (rubella virus) head to trunk to extremities, cervical LADErythema infectiosum (5th dz, Parvo B19) – Red rash on cheeks (slapped cheeks), lacelike rash on trunk/extremitiesRoseola infantosum (6th dz, HHV 6) – rash from trunk to peripheryHand/foot/mouth dz (coxsakievirus type A) – vesicular rash on palms and soles, ulcers on tongue/oral mucosaScarlet fever (Group A strep) – rash with small papules (sandpaper like)
Mistake #1: cards that include too much information waste time, and help minimally with recall
This card is clearly well-intentioned – s/he is studying the viral red rashes of childhood, and wants to try and connect the information, to not fragment the information. Sadly, however, by making a card so long, they will have to repeat a lot of information unnecessarily. For example, let’s say they could miraculously recall all of the information from the first 4, but forgot the causal organism of Scarlet Fever. If they wanted to make sure they could recall that fact, which is important, they would need to repeat the entire card. Second, for anyone who has made a card like this (and I have made lots), whenever it comes up for review, the desire to pass it at all costs so we don’t have to repeat it is quite strong. Thus, we let our standards slip for how well we have to know the information to hit “good” or even “hard.” However, this makes the cards worthless for the exact thing we are using them for , since we will continue to “pass” the card despite not knowing the information solidly.
For cards that are too long, the simplest thing to do is to separate the facts into separate cards
This may go without saying, but the best way to recall this information, while minimizing repeating facts we already know well, is to make separate cards out of the facts. One example would be:
Rubella – Identify the causative agent and describe the rashes in their locations direction of spread.
Rubella virus – head to trunk to extremities, cervical LAD
You could then do the same thing for the other diseases. This makes reviewing the card much faster, improves recall for the few facts (as opposed to cursory recall of many facts) and cuts down on unnecessary repetition of information I already know.
“But don’t you have to make more cards this way?”
Yes, you do, but ultimately, I’ve found that making more cards with less information on each card actually cuts down on the total amount of time spent reviewing your cards.
“How do I know if my cards are generally the right length?”
This may vary person-to-person, but as a rough approximation, it should take you roughly 1 hr to review 100 “old” cards (cards you have reviewed previously). If they are “new” cards, I would expect it to take anywhere between 1-2 minutes per card, as it may take multiple reviews for the information to stick. So, to review 200 old cards, and 35 new cards in a day should take you roughly 2.5-3 hours.
Rule #2: Have a question for which there is a single, unambiguous answer
This encompasses the next two most common mistakes I see people make. The first is illustrated below:
Oral hairy leukoplakia
Mistake #2: Unclear/ambiguous question
This is relatively simple, and most people will catch this eventually, usually after they’ve reviewed the card several times and can’t remember their thought process when they originally wrote it. The solution to this is also relatively clear: simply clarify your question. Something like,
“What is the infectious cause of oral hairy leukoplakia?”
A related mistake is exemplified below:
Hyperventilation would cause what kind of alkalosis?
Mistake #3: Giving away too much information in the question
If I was trying to remember what hyperventilation would cause, I would want to remember not only that it would be a respiratory and not metabolic (I probably could have guessed this), but I would also want to remember that it would cause an alkalosis. While this is a relatively obvious example, it happens all too often: I will write a question, only to find out later when I’m reviewing it that the information I really wanted to recall was already given away in the question itself.
Rule #3: For information you want to recall, test it explicitly in a question
Finally, I wanted to highlight one final mistake that I made constantly when I began to use Anki – a mistake I would like you to avoid. It is a variant of the first mistake (too much information), in which I put a lot of information into the answer and expect to remember it all. This was one of my first cards, and I keep it in my deck as a reminder of how far I’ve come with my cards (and how far ahead you will be with your cards!):
Parainfluenza virus vs. RSV – What are the major syndromes they cause? BONUS: What surface proteins are associated with each?
Parainfluenza – CROUP (laryngotracheobronchitis) – gradual onset, fever, barking cough, inspiratory stridor (wheezing sound), hoarse voice, variable progression. Has mixed neuraminidase and hemagluttinin, so can agglutinate RBCs.
RSV – BRONCHIOLITIS – variable course: tachynpea, grunting & flaring, indrawing & wheezing, cyanosis, apnea in some. NO HEMAGLUTTININ, so do not agglutinate RBCs. Has F (fusion) and G (attachment) proteins – F protein mediates creation of SYNCYTIUM.
Mistake #4: Expecting to know information in the answer that wasn’t explicitly tested
As you can tell, it clearly is too much information. What’s worse, when I made this card, the only place I had the information that croup and bronchiolitis were caused by parainfluenza and RSV was on this card. Do you think I remembered these facts when they counted? Nope. And why? Because I had this misguided thought that if I just saw the information a certain number of times, at the right spaced intervals, I would remember it all.
Remember: the only information you will recall is the information that you specifically ask yourself to recall.
“But wait! What if I didn’t want to test myself actively on this information, but I might find it useful in the future?”
While not really an exception to the above point (you still won’t remember the information if you don’t test yourself on it actively), you CAN leave information for yourself hidden in cards, in case you want to reference it later, but are unsure if you want to test yourself actively on it.
There you have it! Here are the 3 rules for writing basic flashcard-type cards, and the 4 mistakes to avoid:
- Rule #1: Keep the amount of information on a card to a minimum (ideally 1-3 unrelated facts)
- Rule #2: Have a question for which there is a single, unambiguous answer
- Rule #3: For information you want to recall, test it explicitly in a question
- Mistake #1: cards that include too much information waste time, and help minimally with recall
- Mistake #2: Unclear/ambiguous question
- Mistake #3: Giving away too much information in the question
- Mistake #4: Expecting to know information in the answer that wasn’t explicitly tested
What to do next?
- Get started! Download Anki if you haven’t, and make some flashcards!
- When reviewing, note how long your reviews are taking. It should take you roughly 1 hr to review 100 old cards, and 1-2 minutes per new card. If it’s taking longer, re-evaluate your cards: they are likely breaking one of the 3 rules (usually Rule #1: too much information).
- Subscribe to YOUSMLE below, to get some of the highest-yield Anki cards that I used to get a 270 on the USMLE Step 1!
- Check out the Resources page for everything I used and recommend for using your time most efficiently.
- Once you have made basic cards for about a week and feel comfortable with the process, challenge yourself with more advanced card-making skills here.