Whenever you learn something new, you need to refresh your memory at certain points in the future to make sure you don’t forget. For medical students, it’s an impossible task to track how well you know each piece of information you’ve learned and when you need a review. Especially when not all facts are equal - some facts are rapidly forgotten while others can stay in your memory for years.
This is where Firecracker comes in. Firecracker uses 20 different proven learning principles to optimize when, what, and how you study. The most important learning principle is spaced repetition. As the New York Times puts it “You can water a lawn once a week for 90 minutes or three times a week for 30 minutes. Spacing out the watering during the week will keep the lawn greener over time.” Research has shown that repeated testing and active recall of information using a spaced review system can help you learn 10 - 50 times faster than conventional methods.
To create the best possible spaced review system, Firecracker uses an algorithm that has been trained using the quiz data of 100,000 med students who have answered over 148 million questions in Firecracker.
Here’s how it works:
Once you’ve learned a new concept, your ability to correctly recall that concept decreases gradually over time following the Ebbinghaus forgetting curve. Not all forgetting curves are equal though - they depend on how many times you’ve studied a given concept and how well you knew the concept on each review. Firecracker’s algorithm models this forgetting curve for every concept you’ve studied in the app. Our algorithm keeps track of everything you know - it knows what you know better than you do! Once you’re at high risk of forgetting a concept, we’ll recommend that you review it. We use our algorithm to keep you on top of what you’ve learned by prioritizing the concepts that you’re at the highest risk of forgetting.
If you fall behind you’ll have a few more high risk concepts to study, but if you get ahead then you’ll have fewer. Our algorithm takes care of recommending what you need to study by adapting to your performance.
Now compare traditional learning to Firecracker’s optimized spaced learning system. With traditional learning, you review material with arbitrary intervals each review and often end up cramming. With Firecracker’s optimized spaced learning system, your ROST (return on study time) is maximized using a calculated model that prevents you from forgetting what you’ve learned.
How Firecracker’s algorithm was designed:
In order to maximize the efficacy of each of your review sessions, Firecracker’s spaced repetition algorithm was designed to minimize your assigned workload but also ensure that knowledge is never forgotten:
- The intervals between each review should be as long as possible to obtain the minimum frequency of repetitions, and to make the best use of the spacing effect
- Longer and expanding intervals between each review are better for improved recall
- Intervals should be short enough to ensure that the knowledge is never forgotten
- We recommend that you review concepts once you’re on the verge of forgetting them
These two principles used together give you an adaptive recommendation system that will keep you on top of what you’ve learned and minimize the amount of time you need to invest to do so.
And we go even further for you:
When you first start using Firecracker, we’ll ask you your desired exam score and when you’re taking your exam. We’ll use your exam score to adjust our algorithm (if you want a high score you’ll need to review more) and your exam date to recommend a pace at which to study new content on Firecracker.
Our recommendation algorithm does not live in a vacuum. We know that not all concepts are equally important for you. You can assign the relative importance of different content areas by marking them as either “Urgent”, “Current”, or “Past”. Our algorithm will adjust to quickly push “Urgent” concepts to mastery while keeping “Past” concepts on the backburner. In other words, you’ll see “Urgent” questions more often than “Current” questions and you’ll see “Current” questions more than “Past” questions in your daily quizzes.
To read more about how our algorithm works click here. We pride ourselves in having built a superior, evidence based learning tool for our future doctors and we already have strong evidence that Firecracker works.
Data Scientist, Firecracker