Anatomy in medical school, statutes and case law for the bar exam, vocabulary lists for a foreign language. The more serious the study, the more the items to memorize run into the hundreds or thousands. At that scale, a problem appears that almost never comes up when you are memorizing a handful of things: the review process itself breaks down.
"I lost track of how much I've learned." "The material I meant to review piled up so much that I just pretended it wasn't there." This is not a lack of willpower. It is something that happens because of how the system is structured. This article lays out how the breakdown happens and how to run your reviews so that it doesn't.
Why it breaks down: reviews pile up like debt
Every new item you learn creates future reviews. If you learn 20 items today, that generates a scheduled review of those 20 items tomorrow, again a few days later, next week, and next month. In other words, learning something new is placing an order with your future self.
When the following two things combine, the snowball starts rolling.
- The order size is not steady. If you push through 100 items on a motivated day, the wave of reviews for them hits all at once a few days later
- The record exists only in your head. If you track which item to review when from memory, then once the number of items grows, the tracking itself becomes a memorization task and is the first thing to fall apart
The trap is that up to a few dozen items, sheer effort is enough to keep things going. The method that was working stays the same while the number of items keeps growing, and one day it suddenly stops working.
Fix 1: Cap the number of new items per day
The most effective step is to keep the amount of new material you learn each day constant.
Your future review load is determined by how much new material you learned in the past. If you fix new learning at 20 items a day, your daily review load a few weeks later will settle at a roughly constant level. If the amount of new material swings up and down, your review load swings with it, and whatever you couldn't get through on heavy days starts to back up.
Even if it feels slow, a steady pace is faster in total.
The fastest pace is the one that never stops.
Fix 2: When reviews pile up, stop trying to redo everything
Reviews will still pile up. You get sick, work hits a busy season, you don't touch your material for a week. These things happen. The real problem is how you come back. This is where many people fall into one of two patterns: they try to clear the whole backlog and give up, or they start over from the beginning and end up knowing only the early material unusually well.
The key to coming back is triage.
- Sort the backlog of reviews by importance for your exam or for practical use
- Starting with the most important, do only as much as you can manage today
- Spread the rest over the following days. Don't push yourself so hard that you stop all new learning until the backlog is cleared
Even if you have forgotten some items, you are not starting from zero. Material you have learned once is known to take less time to relearn than it did the first time (this is the idea behind the "savings score" in the forgetting curve). A few days of backlog set you back less than you think.
Fix 3: Make each item lighter with recall-based review
The more items you have, the more the review cost per item matters. The classic way efficiency drops here is spending time rereading.
The rule is that reviewing means recalling without looking, not looking to confirm (this is the testing effect, also called retrieval practice; see the article on spaced repetition for details). A recall-based review takes a few seconds to a few dozen seconds per item, and it also produces better retention than rereading.
- Items you can recall: you're done. Move on
- Items you can't recall: only then go back to the material
Not "everything, equally carefully," but "spend time only on what you can't recall." That is the best way to allocate your time when you have a large number of items.
Summary
- When review management breaks down, it is a structural problem, not a willpower problem. Learning something new is placing an order for future reviews
- Keep the amount of new material constant each day, and your review load settles at a steady level and keeps running
- When reviews pile up, don't redo everything. Triage. Relearning goes faster than the first time, so you can catch up
- Review by recalling, not rereading. Spend time only on the items you can't recall