I talked earlier about spaced repetition and SRS software, and I said then that I would write a bit about best practices for making spaced repetition part of your study, and sticking with it for long enough that it will really make its impact obvious.
Trust the machine
One of the most common reasons for people to fail with and stop using spaced repetition is inconsistency. As I mentioned earlier, the key to SRS is that the system can calculate the latest possible repetition, thereby greatly reducing the total workload. While this calculation isn’t perfect — the algorithm is, after all, just a best guess approximation of how your memory actually works — its pretty good. What this means, though, is that if you miss your scheduled review date, the likelihood of forgetting the fact increases rather quickly.
Most SRS systems are, I believe, conservative, opting for slightly more reviews in order to reduce the odds of forgetting something. However, if you miss your review date by a wide margin (relative to the age of the fact — new items are forgotten much faster than old items), there’s a pretty good chance that you’ll forget what you’ve learned.
For best results, you need to review every day. You don’t need to add new material every day, though, so if you’re feeling overloaded you can hold off on adding new items into your reviews. Your scheduled reviews, though, should get done unless absolutely impossible. After all, forgetting has no pause button.
Don’t review too much at once
No matter how hard you try, your attention span is limited. According to Wikipedia:
Estimates for the length of human attention span are highly variable and range from 3 to 5 minutes per year of age in young children, to a maximum of around 20 minutes in adults. Nevertheless, the average movie lasts approximately two hours, and most adults can follow the plot with only minimal lapses of attention.
I suspect the exact length of time you can study without your mind wandering varies between individuals. I strongly recommend that you figure out what your attention span is and respect it. I find for me that about 10 minutes is perfect, after which I start to get bored. I break my SRS repetitions into as many 10 minute segments per day as needed to complete my reviews for the day, ensuring that I get what I need to get done without getting too bored.
Review when you’re most alert and ready to learn
Different people are at their mental best at different times. I’m an early bird, and so I do most of my studying before 7am. I have friends whose brains don’t even start working before 11am, and do their best work at night. To each their own.
If you try to study when you’re not alert, though, you’re simply wasting your time. Don’t do it — you’re not getting any benefit from it, and when you realize your failing all of your reviews anyway it will just hurt your motivation. Get some sleep instead, and tackle your learning when you’re awake.
Don’t try to memorize what you don’t understand
The title basically says it all. The difference between reviewing what you’ve already learned — good! — and memorizing what you don’t know — bad! — may not seem particularly large, but the results will be. I’ve tried dumping vocabulary lists into Anki and memorizing those, and while I generally do memorize the words eventually, its painful and demotivating. If, for instance, you don’t know how to use a particular word in a sentence, then memorizing its meaning is going to be hard (and of dubious value).
What works best for me is to gather new things from other contexts — books, movies, TV, conversations, etc. — and then at the end of the day convert the notes from my notebook into Anki items to review.
Context is everything
Well, maybe not everything, but it is certainly very important. As I just wrote above, I like to take new information from other contexts and enter it into Anki. Ideally, that information was recorded along with the context in which I first found it. If not, I try my best to manufacture context for the purpose of creating the card.
I also like to approach information from several different directions. For instance, if I learn a new word, I might create four or five different cards, all containing that word. Examples include:
- The word itself. The simplest card, with the word on one side and its definition in either English or my target language on the other.
- Common phrases. If the word is commonly used along with other words, I try to learn those colocations as phrases, as I find if I review them separately they enter my active vocabulary more quickly.
- Sentences. For every new word I enter, I enter at least one sentence using that word. I get these sentences from a variety of sources. If I have access to the original context (i.e., from a book or blog post) I use that sentence. Many dictionaries (including most learner’s dictionaries) include example sentences for most or all of their entries. Also, Google News can be a good friend here, letting you find recent news stories that contain the word you’re learning.
Stick to the minimum information principle
Put simply, each card should test a single thing. The more complex your cards are, and the more information you’re being asked to recall on each review, the less likely you are to remember any of it. It is much better to create several simple cards than to try to review everything about a particular topic on a single card.
Whew!
And there you have it, my tips for better SRSing. Do you have any of your own? Let me know in the comments or by e-mail.
Related posts:
- Spaced repetition for beginners
- Self-regulating Anki settings
- Avoiding SRS burnout and repetition hell
- Thoughts on SRS content selection
- You’re doing it wrong
{ 2 comments… read them below or add one }
Great tips! I find myself rereading the article periodically for small nuggets of advice. For example, I thought I understood the definition of some new words and the example given in the sentence. But in the back of my mind, I knew I didn’t truly understand or know what context to use the new word even with the examples given in the book. By after rereading your article, I used Google news to find additional real world usage examples that helped me to fully understand the new term.
Also, I find more often than not that my mind recalls the context or usage example prior to recalling the definition. The usage examples usually pops rather quickly and then like magic, the definition just pops from my memory. I wonder if this is a good thing??? I suspect it would be cause I memorized definitions previously without any contexts but it definitely is painful to recall it in the future cause of the lack of some contexts.
Thanks agian!
I suspect it’s a good thing (the same thing happens to me). It’s natural, I think. When I think about a word in English, I don’t normally come up with dictionary definitions, but rather usage examples and synonyms. I think that’s basically what’s happening here, too.