Back to blog

How to Study for an Essay Exam with Flashcards (Card the Argument)

Do not memorize a finished essay. Card the claims, the evidence under each claim, and the counterargument, then practice five minute outlines from them.

You can study for an essay exam with flashcards, but not by putting essays on them. What goes on the cards is the raw material an essay is built from: the claims you might make, the specific evidence that supports each claim, and the strongest objection to it. Then you practice the building part separately, by writing five minute outlines to prompts you have not seen.

The approach that fails is the popular one. You guess the question, you write a full essay at home, you memorize it paragraph by paragraph, and then you walk in hoping the prompt matches. When it matches, great. When the prompt comes at the topic from a slightly different angle, you are stuck with eight hundred words that answer a question nobody asked, and you either write it anyway or you improvise with whatever pieces you can pull loose.

I work on Nebulearn, a flashcard app, so factor that in. I will also say up front that the cards are maybe half of the preparation here, and the outlining practice is the other half.

Why the memorized essay breaks

It is worth understanding exactly how the memorized essay fails, because the fix follows from it.

When you memorize an essay, you memorize it in order. Sentence one cues sentence two, which cues sentence three. That is a chain, and a chain is only useful if you start at the beginning and go straight to the end. It is the same thing that happens when you only know your flashcards in the order they come in, which I wrote about in the context dependency problem.

Now you are in the exam. You prepared "discuss the causes of the First World War," and the prompt says "to what extent was the alliance system responsible for the outbreak of war in 1914?" Those are close, but they are different questions. The second one wants an argument about one cause weighed against the others. Your essay goes through four causes in sequence and gives each of them a paragraph. You read the prompt and you think, "okay, this is basically my essay." It is not. A marker reading your answer sees a list of causes and no judgment about extent, which is the one thing the prompt asked for.

What you needed was the same knowledge in a different arrangement. You needed to be able to pull out the alliance paragraph, lead with it, and then use the other three as the "to what extent" comparison. If your knowledge is stored as a chain, you cannot rearrange it. If it is stored as separate pieces, each of which you can recall on its own, you can build whatever the prompt asks for.

That is what the cards are for. Each one is a piece you can pick up independently.

The three kinds of card an essay is made of

An essay answer, stripped down, is a position, a handful of claims that support it, evidence for each claim, and some acknowledgment of the other side. So those are the cards.

Claim cards. The front is a theme or a question the course keeps returning to. The back is a one sentence position you could defend. One sentence, and it has to be arguable.

Front: What is one defensible position on how much the alliance system mattered in 1914?

Back: The alliances did not cause the war, but they turned a regional crisis into a continental one by making each power's decision depend on its partners.

Evidence cards. This is where most of your deck should be. The front is a claim, and the back is one specific piece of evidence for it: an event with its date, a quotation, a statistic, a named scholar's argument, a scene from the text. One piece per card.

Front: One specific piece of evidence that the alliance system widened the 1914 crisis instead of starting it.

Back: Germany's backing of Austria-Hungary in early July 1914, the so-called blank cheque. The original dispute was between Austria-Hungary and Serbia.

Counter cards. The front is your claim, and the back is the best objection to it, plus a line on how you would answer. Markers reward the essay that deals with the other side, and it is the part students most often leave out, because it was never something they sat down and learned.

Front: Strongest objection to "the alliances only widened the war."

Back: The alliances were defensive and loosely worded, and Italy did not follow its partners in 1914, so they were not automatic. Response: they shaped expectations and planning even where they did not bind.

Each of those is small enough to grade honestly in a few seconds. That matters, because the natural essay card is "discuss the causes of the war" with a full outline on the back, and that card is a five line answer you will always rate yourself generously on. The rule from atomic encoding applies with full force here. If the back has more than one idea, split it.

How many cards per theme

A fair question is how this stays manageable, because "every piece of evidence in the course" sounds like a thousand cards.

It is not, because you are not carding the course. You are carding what you would actually use. Start from the likely themes. Most essay courses have somewhere between five and ten big questions that the lectures keep circling, and the syllabus or the old exams will tell you what they are. For each theme:

Card typeHow many per themeNotes
Claim2 or 3Different positions, so you can pick the one the prompt favours
Evidence3 or 4 per claimSpecific: a name, a date, a line, a number
Counter1 per claimThe best objection, not a weak one
Key termsAs neededAny concept you must define precisely to be taken seriously

That works out to roughly fifteen to twenty cards a theme, so a hundred to a hundred and fifty for a typical course. That is a reasonable deck for a few weeks of review. If you are far above that, you are probably carding lecture content you would never put in an essay. The filter from the 50-slide lecture triage works well here: for each candidate card, ask whether you can picture the sentence in an essay where you would use it. If you cannot, cut it.

Evidence cards are also where quotations go, if your subject needs them. Keep quotations short. A six word phrase you can reproduce exactly is worth more than a three line passage you will paraphrase badly. Put the phrase on the back, and on the front put the point it supports, not "quote from chapter four," because in the exam you will be going from the point to the quote and never the other way round.

The other half: five minute outlines

The cards give you the pieces. They do not give you practice putting the pieces together under time pressure, and that is a separate skill. If you only do the cards, you will walk in with a head full of well learned fragments and spend the first fifteen minutes of each essay working out how to arrange them.

So a few times a week, you do this. Take a prompt you have not answered before. Old exam papers are best. If you run out, take a prompt you have done and change the command word: turn "discuss" into "to what extent," or "compare" into "which was more important." Set a timer for five minutes. On a blank page, write:

  1. Your position, in one sentence, answering the actual prompt.
  2. Three or four claims that support it, one line each.
  3. Under each claim, the evidence you would use. Just a few words, enough to show you have it.
  4. Where the counterargument goes and what you will say about it.

Then stop. Do not write the essay. Check the outline against your cards. Which evidence did you forget you had? Which claim had nothing under it? That tells you which cards are weak, and it tells you far more accurately than the review session does, because this is the real retrieval the exam will ask for.

This is blank sheet retrieval pointed at essays. Five minutes is on purpose. In the exam you will spend about that long planning each essay, and an outline you built in five minutes is one you can then just write out, which is where the marks are.

Every week or so, pick one outline and actually write the full essay, timed, by hand if the exam is by hand. Handwriting speed and stamina are a real limit, and you do not want to find out in the room that your hand cramps at minute forty.

Where the cards should come from

The best source for claim and evidence cards is your reading, at the moment you are doing it. When you hit a sentence in an article that makes you think, "that is a good point, I could use that," that is a card. The trouble is the usual one: you think, "I will come back and note that down later," and later it is a yellow highlight with no memory of why you highlighted it.

I wrote about the general version of this in the evidence-claim PDF extraction workflow. The short version is that a highlight is not a card. A card needs the claim on one side and the evidence on the other, and you only know which claim a piece of evidence supports while you are reading the paragraph it sits in. So you card it then, in one sentence, and move on.

Secondary sources matter here too. If your course expects you to engage with scholars, make a card for each major one: the front is the name and the debate, the back is their position in one sentence. Being able to write "as one line of scholarship argues" and attach a name is a cheap way to make an answer look like it belongs in the conversation.

Where other approaches are better

Flashcards are not the best tool for every part of this, and some courses barely need them.

If the exam is an unseen passage or source and you are asked to analyze it on the spot, there is very little to memorize. The preparation is practicing analysis on passages you have not seen, and a deck would mostly be a way of feeling busy. A small set of cards for technical terms is all I would make.

If your knowledge of the subject lives in the connections between ideas, a notes tool that links things together, like Obsidian or RemNote, may suit you better than a stack of separate cards. Seeing that three readings all bear on the same theme is exactly what a linked note system is good at, and it is a real weakness of flashcards that each card stands alone.

If you already use Anki and have a routine, there is nothing in this method that needs a different app. Tags for themes, and the three card types above as plain front and back cards, will do everything described here. And for the outlining half, the best tool is paper and a timer.

When it still goes wrong

Two things tend to go wrong even with good cards and regular outlines.

The first is answering the theme instead of the question. You see a word in the prompt that matches one of your themes, and all the cards for that theme come flooding back, and you write down everything you know. It will be accurate and it will not answer the question. The fix is in the outline step: sentence one has to use the prompt's own command word. If the prompt says "to what extent," your first sentence should contain something like "largely," "partly" or "only in the sense that." If it does not, you are about to write a description.

The second is evidence going vague. Three weeks out you could give the date and the name. The night before, it has faded to "around that time, some official said something like." That is ordinary forgetting, and it is the thing a scheduler is for. Evidence cards need the spacing more than any other type, because specific details fade fastest, which is the point I made in why you cannot remember what you read.

The split in this article is between the essay and the pieces of the essay. The finished essay is a chain you can only run in one direction. The pieces, kept on separate cards, can be put together to fit whatever shows up. That is how I would set it up in Nebulearn: read the source in the side by side PDF view, write the claim and the evidence while the paragraph is still in front of you, and let FSRS, which is on by default, keep the specifics from fading before the exam.

Essay ExamsFlashcardsHumanitiesExam Prep