On a Newsweek article titled “What’s Your Name Again?” (April 27, 2009, page 57), Harvard psychologist Dr. Aaron Philip Nelson answers questions about memory. Here is an excerpt:
Question: I am a college student. What’s the best way to retain large amounts of information for the long term?
Answer: The single best way to retain large amounts of information over time is to rehearse it periodically. Revisiting the material induces the brain to consolidate the information, thereby strengthening the neural network in the brain that contains the information as a pattern of electrical activity. For long-term memory, occasional rehearsal is much more effective than one-time cramming.
ISORM is a technology based on that principle. It maximizes learning, with the least possible effort. For the list of research papers supporting it please go to the REFERENCES section below.
ISORM is a knowledge-delivery system that dynamically adjusts to your speed of learning. ISORM detects the Objects you have difficulty learning and schedules reviews on them more frequently. That is, the program learns from your mistakes in order to schedule review quizzes.
The ISORM method can be quickly understood with an example.
Suppose you are learning Spanish and you are right now studying the word "éxito" (success) and then take a quiz and pass it. In this case, the quiz is testing your short-term memory. The word "éxito" will likely fade from your memory as the hours pass by.
How do you make sure that you will remember "éxito" in, say, ten years?
The answer is: learning with ISORM. Continuing with the example, ISORM schedules review quizzes intelligently until you can remember, without any help, the meaning of the word "éxito" thirty days after having last seen it.
If you remember "éxito" without any help, after thirty or more days, the probability that "éxito" is securely stored in your long-term memory is close to 100%.
Long-term human retention of studied material increases when the material is reviewed or tested at ever-increasing intervals. The process follows this principle by establishing a system of points and testing intervals (measured in days) by which the student’s memory retention of a given item is automatically tested. The process increases or decreases the intervals depending on the ability of the student to remember said item. The automatic testing of said item stops when a given number of days (ideally thirty or more) have lapsed since the last test, and the student can remember said item. At this point there is a high probability that the item is finally stored in the student’s long-term memory.
The process of scheduling the date of subsequent tests, and the execution and grading of said tests for each item studied, could be done by a human-being using paper and pencil, but it would necessitate the dedication of an enormous amount of time, which makes it impractical. A computer, however, can perform the task quickly and automatically.
In order to improve the retention of information in a student’s long-term memory, tests are traditionally used. The problem with traditional tests is that students usually prepare shortly before in a cramming session and the material is quickly forgotten. Final or reviewing exams do not solve this problem because, again, students prepare only shortly before the exam.
The general idea of ISORM is that the software gives the student the opportunity of forgetting the studied material by letting several days pass-by between tests, increasing the interval as the item is becoming fixed in the student’s long-term memory.
The probability that a given item has become fixed in a student’s long-term memory is proportional to the number of days that have lapsed since the last test of said item, provided that the student correctly answers the test on said item. For example, if the item consists of the German noun “das Gehirn” (brain) and a student remembers it thirty days after its last test, the probability of that item being fixed in his or her long-term memory is much higher than if the correct answer was produced five days after its last test.
The automatic testing of a given item stops when the student can correctly recall it after a number of days (ideally more than 30) have lapsed since the last test on said item. When this occurs, the item is considered learned, and is removed from the automatic-testing process.
The process is closely related to its computer application because it cannot be practically applied without a computer. That is, a human being without the aid of a computer would need to spend so much time carrying out the process that it would render it useless. For example, if a student is learning 3000 German words, and on average it takes him ten tests to learn and item, and it takes one minute to execute the process for a typical item (producing the test, grading it, scheduling the date of next test and recording the item in the list of Objects to be studied on said date), then it would require 30000 minutes, or 500 hours to perform the process without the computer, whereas the computer would perform the task virtually instantaneously.
The process accommodates slow and fast learners. A fast learner would make less mistakes on the tests and would exit the process quicker than a slow learner.