Supplemental D - Research & Science on Memory, Spaced Repetition, & Active Recall
Background reading: 📝 Lesson 3 - What is Anki and why should I use it?
-
We forget quickly
-
Spaced repetition
-
Active recall
-
Students choose easier but less effective methods
-
Real world implications
-
Anki
The science behind spaced repetition starts with a simple problem: we tend to forget quickly.
In the late 1800s, Hermann Ebbinghaus conducted the classic study of memory, where using lists of ‘nonsense syllables’ described what is now known as the ‘forgetting curve’. His research revealed that memory tends to drop off steeply soon after learning, and slowly decline over time.
A century later, Murre and Dros in 2015, replicated Ebbinghaus’s protocol in which they found similar results, confirming his initial findings. This supported the broader idea that we tend to forget information over time in a similar and predictable fashion, especially when paired with no recall of the information, and this can be used to the advantage of algorithms that can predict when we will forget information, and when we will need to study it again to effectively retain it.
- Spaced repetition
This is where the concept of spaced repetition plays a vital role. Instead of reviewing material once and hoping it sticks, spaced repetition spreads review sessions over time. In 2006, Cepeda et al. published a meta-analysis that concluded the benefit of spacing out study over time, rather than “massed” learning, is superior. Donovan and Radosevich, in 1999 quantified this benefit; their meta-analysis of 63 studies showed that individuals using spaced practice outperform those using massed practice.
Later research refined the idea of distributed practice. We know that spacing reviews apart is effective, but another question arose. How do we space them? In 2008, Cepeda and colleagues found that the best interval between study sessions depends on how long the learner needs to retain the information. In simple terms, very short gaps are often inefficient, long gaps lead to too much forgetting, and an intermediate gap tends to work best. This is one reason spaced repetition software, such as Anki, is useful: reviews are scheduled near the point where you are most likely to forget the information, rather than on a fixed schedule.
- Active recall
There is another piece to the puzzle, active recall. Active recall, also referred to as retrieval practice or the testing effect, refers to the act of actively retrieving information out of memory rather than passive rereading. In 2006, Roediger and Karpicke wanted to see whether re-testing oneself was effective because you are “seeing” the information again, a form of restudying, or there was inherent value in testing oneself on retention. They found that rereading could improve immediate performance, but active recall produced better retention days later. In 2014, Rowland’s meta-analysis further supported the testing effect, revealing that retrieval practice improves retention compared to simply restudying. Importantly, stronger benefit was seen for recall-based tests than recognition based-tests; more effortful retrieval tends to produce stronger long-term learning.
- Students choose easier but less effective methods
This also helps explain why many students study inefficiently. Students often rely on strategies like rereading, highlighting, or reviewing notes because these methods feel productive. But research on learning techniques consistently ranks practice testing and distributed practice among the most useful strategies for long-term learning. In other words, the most effective methods are often the ones that feel less comfortable in the moment.
Bjork, in 1994, described this broader idea as “desirable difficulty”, learning conditions that slow performance in the short term but improve long term retention and transfer. The difficulty of those methods is part of what makes the learning more durable. Bjork, Dunlosky, and Kornell later emphasized that learners are often poor judges of their own learning. Strategies, such as rereading or reviewing notes, create confidence without durable recall. Strategies that feel harder, such as spaced repetition or self-testing, may feel less productive in the moment but produce better long term learning.
Dunlosky and colleagues reached similar conclusions when reviewing common study techniques. Popular methods included rereading and highlighting, but do not consistently improve long term performance, while practice testing and distributed practice were rated as strategies with high utility.
This further highlights an important aspect of learning. Students often are expected to learn and memorize, but are not taught how to do so, how to memorize effectively, or what methods are most optimal. Instead of placing the blame on the student, we should be asking whether they were ever taught the right strategies in the first place. Too often, the blame is placed on the individual, when the real issue may be a system that expects people to perform without properly teaching them how to learn.
- Real world implications
These findings translate in the real world and the classroom, as described in a systematic review by Agarwal and colleagues in 2021, where retrieval practice was beneficial across a host of variables, such as education levels and content areas. This is especially relevant in medicine, a field where medical students, residents, and physicians are expected to retain a large volume of information and retrieve it under pressure, correctly, and in a timely manner. Recent research in the medical education space supports the use of spaced repetition in these specific settings. In 2026, Maye and Hurley, conducted a meta-analysis on spaced repetition, compiling 21,415 learners and found that spaced repetition improved performance on objective tests compared with standard studying techniques. Furthermore, a 2023 cohort-control study by Gilbert et al., also found that Anki use was associated with higher exam scores among medical students. In practicing physicians, a large prospective cohort study by Price et al in 2024 found that spaced repetition improved both learning and, subsequently, knowledge transfer.
Taken together, the aforementioned studies suggest that spaced repetition is useful not only for exam preparation, but for maintaining and transferring knowledge over time.
- Anki
Anki, the free, open-source, application created by Damien Elmes in 2006 solves this issue and combines these principles into a practical workflow. Each flashcard prompts active recall, and the scheduler spaces future reviews based on the user’s previous performance. Older algorithms, such as SuperMemo-2, helped make this possible, with the future pointing to newer and more sophisticated algorithms, such as the free spaced repetition scheduler (FSRS) developed by Junyao (Jarrett) Ye, aiming to better estimate when a learner is likely to forget and adjust future reviews accordingly, using factors such as retrievability and stability.
Anki makes evidence-based learning and memorization a priority and necessity. Throughout academia, we are always taught what we should learn and memorize, but never how we should do so. Anki turns two of the strongest findings in memory research, spaced repetition and active recall, into an effective daily system for lifelong learning. For medical learners, the challenge is not just establishing the information into memory, but keeping it available, easy for retrieval at a moment’s notice, for years to come. Anki is the solution that makes it possible.
