Are Flashcards Effective? Keep the Cards You Already Know
- Flashcards are effective because they make you recall an answer before you see it.
- Students who dropped a word once they got it right recalled 33 to 36% a week later; those who kept recalling it, about 80%.
- Seeing a learned card again did nothing for next week; recalling it again made the whole difference.
- One big stack beat four small ones for 90% of learners, yet 72% believed the small stacks had worked better.
- Keep the cards you know in the stack, and recall them a few more times at growing gaps.
Flashcards are effective. But the benefit does not come from making the cards. It comes from recalling the answer before you flip. So when you set aside a card the moment you get it right, the benefit goes with it. Students who dropped words once they got them right recalled 33 to 36% of them a week later; students who kept recalling them recalled about 80%.
Are flashcards effective?
Yes. Flashcards make it easy to test yourself, and testing yourself is one of the best supported ways to study.
Dunlosky and colleagues (2013) reviewed ten study techniques and rated two as high utility: practice testing and distributed practice (spreading study out over time). Rereading, highlighting and summarizing were rated low. With a flashcard you look at the front, recall the answer and check the back. Used as intended, it is practice testing.
The trouble is in how people use them. Two habits come naturally with flashcards, and both cut into the benefit: setting aside the cards you got right, and splitting the deck into small stacks you finish one at a time.
Should you drop flashcards you already know?
No. Getting a card right once does not mean you have learned it. You need to recall it again after that for it to last a week.
Karpicke and Roediger (2008, Science) had college students learn 40 Swahili-English word pairs (for example, mashua-boat). Everyone studied the list once and was tested once in the same way. After that, what happened to a word once a student got it right was split four ways:
- Words got right were still studied and still tested.
- Words got right were no longer studied, but stayed in every test.
- Words got right were still studied, but dropped from tests.
- Words got right were dropped from both. This is what many popular study guides recommend.
During learning, the four groups did the same. By the end, nearly everyone could recall all 40 words. The students' predictions for a test a week later were alike too, 20 to 22 words out of 40 in every group. The actual results a week later were not alike at all.
Dropping learned words from testing cut next week's recall by more than half
- Predicted
- Actual a week later
See the numbers
| Predicted | Actual a week later | |
|---|---|---|
| Study + recall | 52% | 80% |
| Recall only | 51% | 80% |
| Study only | 55% | 36% |
| Neither | 50.8% | 33% |
The two groups that kept recalling words they had already got right recalled about 80%. The two that dropped them from testing recalled 36% and 33%. The scores did not even overlap: the best student who dropped words scored 60%, and the weakest student who kept recalling scored 63%.
Two things stand out. First, seeing a learned word again made no difference a week later. The group that kept studying every word in every round, but stopped being tested on the ones it knew, still landed at 36%. Only recalling made the difference. Second, the students had no idea. Every group predicted around 50%.
In one line: dropping the cards you got right saves time now and costs you the memory next week.
Keeping learned cards in the rotation does take more trials. The paper's Table 1 shows how many.
| What happened to words got right | Trials during learning (study and test) | Recalled a week later |
|---|---|---|
| Kept studying and recalling | 320 | about 80% |
| Stopped studying, kept recalling | 236.8 | about 80% |
| Kept studying, stopped recalling | 243.0 | 36% |
| Dropped from both | 154.8 | 33% |
The group that stopped studying but kept recalling did about 80 fewer trials than the busiest group and scored the same. So you do not need to reread the back of a card you know every time. You just need to keep recalling it from the front.
Kornell and Bjork (2008) found the same direction across four experiments. When students were allowed to set aside cards they thought they knew, learning got slightly but consistently worse. The authors wrote that dropping rests on two shaky foundations: students' judgments of what they know, and the value they place on further study.
Should you study flashcards in small stacks?
One big stack works better. The gap before a card comes back is longer, so each recall takes more effort.
Split 100 cards into four stacks of 25 and you pass 24 other cards before a card comes back. Run all 100 as one stack and you pass 99. Kornell (2009) tested this with a web-based study program in which people learned GRE-type word pairs. Across three experiments, studying one large stack beat studying four smaller stacks, and it also beat cramming on the last day before the test.
The big stack worked better for 9 in 10 learners, yet 72% believed the small stacks did
See the numbers
| Share of participants | |
|---|---|
| Big stack actually worked better | 90% |
| Believed small stacks worked better | 72% |
It is easy to see why small stacks feel better. A card you just saw comes back soon, so you get it right easily, and getting it right easily feels like learning. But what comes back easily fades just as fast. How long to wait between reviews is in our note on the spaced repetition schedule.
How many times do you need to get a card right?
One study suggests about three correct recalls at first, then about three more on later, widely spaced days.
Rawson and Dunlosky (2011) had 533 college students learn conceptual material through retrieval practice. In the first session they practiced each item until they had recalled it correctly 1 to 4 times. Then, in 1 to 5 later sessions, they relearned each item to one more correct recall. Memory was measured 1 to 4 months later. Getting items right more times at first helped a lot before any relearning, but that advantage shrank as relearning sessions were added. Relearning greatly improved long-term retention at little extra cost in practice. The authors' advice: practice to 3 correct recalls, then relearn 3 times at widely spaced intervals.
Students know half of this. Wissman and colleagues (2012) surveyed 374 undergraduates about how they use flashcards. Students understood the benefit of practicing until they got items right more times, but they did not usually leave long gaps between practice, and did not understand why that helps.
How should you use flashcards, starting today?
Keep the cards you know in the deck, keep the deck big, and come back to it on later days. You can do all of this with paper cards.
- Say or write the answer before you flip. If you flip at "I think I know this", no recall happened.
- Do not set aside a card you got right; move it to the back of the stack. Recall it again when it comes up. You do not need to reread the back for long.
- Run one stack instead of several small ones. If it is too many, put in only what you can get through today and add the rest tomorrow.
- Move a card you have got right about three times into a "later" box. Pull that box out in two days, a week and a few weeks, and recall them again. Any you miss go back into the daily stack.
A spaced repetition app such as Anki does step 4 for you. It does not throw away cards you get right; it pushes each one's next showing further out. If you use an app, the main thing is to stop deleting or suspending cards because "I know this one." For why cards written from memory beat cards copied from the text, see Is note-taking a waste of time?
In Aloom
Aloom has no button for setting a card aside. You ask AI, and the answers you keep stay in your knowledge graph as nodes. Each day, 1, 3, or 5 of them come back as questions, and you recall before you see the answer. The ones you get right stay in the graph and come back again after a longer wait. Keeping the cards you know and recalling them again, from this note, is the daily Recall. The more you recall a node, the darker its ink. Take a look at Aloom
FAQ
Is Anki worth it?
What Anki does, making you recall first and showing cards you got right again at growing intervals, is what the studies in this note found works. These studies did not test Anki itself, though. The benefit comes from the daily habit of recalling, not from the app.
Are flashcards effective for language learning?
Yes. The experiments by Karpicke and Roediger and by Kornell were both about learning words (Swahili-English pairs, GRE-type English vocabulary). Flashcards suit anything you can ask and answer as a pair, like a word and its meaning.
Do flashcards work for understanding, or only memorizing?
The studies here measured recall of words and, in Rawson and Dunlosky, of concepts. They do not show that flashcards build deeper understanding on their own. For that, recall the idea in your own words, and check it against the source.
References
- Karpicke, J. D., & Roediger, H. L. (2008). The critical importance of retrieval for learning. Science, 319(5865), 966–968.
- Kornell, N. (2009). Optimising learning using flashcards: Spacing is more effective than cramming. Applied Cognitive Psychology, 23, 1297–1317.
- Kornell, N., & Bjork, R. A. (2008). Optimising self-regulated study: The benefits and costs of dropping flashcards. Memory, 16(2), 125–136.
- Wissman, K. T., Rawson, K. A., & Pyc, M. A. (2012). How and when do students use flashcards? Memory, 20(6), 568–579.
- Rawson, K. A., & Dunlosky, J. (2011). Optimizing schedules of retrieval practice for durable and efficient learning: How much is enough? Journal of Experimental Psychology: General, 140(3), 283–302.
- Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4–58.