How does the human memory work?
Much scientific research has been undertaken to understand the human brain. Herman Ebbinghaus (1885, http://en.wikipedia.org/wiki/Hermann_Ebbinghaus) found some scientific evidence for the learning curve, also called the forgetting curve. If you study your learning material once, but then fail to maintain it, your knowledge will fade away. Initially this will happen very quickly, but after a few repetitions the process is slowed.
If you watch a video today, or read an article or textbook and subsequently not do anything with it, a week later 80% to 90% of the content will have been forgotten.
The power of repetition
Repitition of learning material turns out to be the only real remedy against forgetting. Paul Pimsleur (1967, http://en.wikipedia.org/wiki/Paul_Pimsleur) noted that the best moment to repeat is when the chance of still remembering the material is at least 60%.
Pimsleur also determined that if items are remembered correctly after repetition, the repetition interval may be decreased. This eventually leads to a much more durable result. This principle was later confirmed by research done by Thomas Landauer and Rober Bjork (1978).
German psychologist Sebastian Leitner (1972, http://en.wikipedia.org/wiki/Sebastian_Leitner) was the first to develop a practical application of the theory of repetition with increasing intervals. Leitner developed the “hand computer”.
In this sytem, items that are mastered correctly are repeated less frequently than items for which incorrect answers were given.
Leitner has scientifically proven that learning material is absorbed more quickly and more durable if the material is presented in “manageable pieces”. If as an example one hundred English words are to learned, this is best achieved by working in blocks of ten words each.
Aside from Leitner’ “hand computer”, Leitner noted that people remember better if learning material is studied in varying physical locations. Later this was scientifically confirmed to be caused by the way the human brain couples background sounds (think for instance of a ticking clock) to making the connection between a known element (a question or math problem) and an unknown element (the answer or the solution).
If the ticking of the clock is no longer present, the brain may have trouble reproducing the connection between the known and unknown elements. This phenomenon is sometimes apparent as a “black out” during a test or exam.
In learning phsycology, the place and time independent learning and clever repetition of learning elements through time is called the “Leitner System”. Drillster has digitized and improved on the Leitner system. Modern technology makes it possible to further enhance the sytem described above. This lead to the Drillster algorithm.
How does Drillster work?
For each individual user, Drillster records:
- Which entries an individual user has seen, and has had a chance to memorize
- Which questions have been answered correctly or incorrectly
- When parts of the material have last been studied.
Using this information Drillster is able to determine the optimal sequence and frequency of the items which are to be mastered. This allows users to learn more is less time, and to retain the knowledge longer.
Elements of the Leitner system that have been incorporated in Drillster:
- Elements answered incorrectly are repeated more often than elements for which a correct answer was given;
- Drillster starts introducing new elements after having trained a small number of items. Drillster teaches and tests knowledge in manageable pieces;
- Drillster is “cloud learning”, which promotes time and place independent learning. The learning material can be absorbed on any desktop, laptop, smartphone or tablet. Varying the means of learning further enhances knowledge retention;
- If a user has not seen a particular drill for some time, Drillster will check whether the user is still proficient. If it turns out that that is not the case, learning material is repeated.
- Drillster sends out notifications when proficiency is still reasonable. This makes brushing up relatively quickly and efficient. After all, it avoid users from forgetting the material altogether.
Other elements of the Drillster system that have a scientific foundation:
- Drillster provides direct feedback for correct and incorrect answers. The human brain is able to process information much better with direct feedback, compared to the delayed feedback that a conventional test provides;
- Using all of the information that Drillster collects during practicing, a very accurate depiction of the true proficiency can be made. Research has shown that Drillster has very good predictive power: the score calculated by Drillster on the evening before a test or exam is very closely related to the eventual result for that test or exam;
- Also, using the collected data, a calculation is made to determine how quickly the absorbed knowledge will disappear. This is done using the learning curve devised by Herman Ebbinghaus. Both individual users and teachers or managers get an accurate insight into the future proficiency. This information can be put to good use in efficiently planning the various learning activities. Learning material is only repeated if it turns out that that is actually needed. No other memorization tools provide this clever mechanism;
- If the brain is rewarded directly for memorizing information, the brain will work harder to memorize things correctly. Actually seeing the proficiency score increase while practicing with Drillster encourages this positive learning process;
Leitner determined that by smart learning it is possible to achieve a 20% to 40% time savings, just by cleverly repeating the relevant items. He also proved that learning material can be kept top-of-mind through time by repetition just before forgetting sets in.
Research with Drillster users has shown that a 30% to 40% time savings could be achieved. Students who had used Drillster were able to achieve 10% to 15% higher test scores.
Drillster has digitized and further enhanced the Leitner system. Drillster’ unique way of calculating a user’s true current and future proficiency is seen as a very valuable didactic asset. The scoring system devised by Drillster has a proven predictive value. The proficiency score indicated by Drillster can be accurately correlated to the result of a test or exam taken at that moment.