Memory maintenance Graphic

The boring truth about never forgetting

by Matthew Leitch, 30 July 2003

Boring but fantastic

The single most effective technique for not forgetting things you have learned is simply to remind yourself from time to time. How dull. But do this efficiently, as described below, and you can increase your learning dramatically, starting today. Either aim for better results or enjoy the spare time you will have.

This technique is especially suitable if you are studying for exams. It was the stunning results I achieved with this technique when studying for school exams many years ago that started my personal interest in memory and psychology.

The quest for efficiency

Remind yourself from time to time. But when and how? The following sections discuss the alternative approaches and when to use them.


The earliest reference to systematic reviews that I have found is in ‘The psychology of study’ by Professor C A Mace, which was first published in 1932. Professor Mace advised readers to review what they had learned after one day, then again after a further 2 days, then after another 4 days, 8 days, and so on.

Tony Buzan, in ‘Use Your Head’, advised a different schedule: after 10 minutes, after another day, a further week, a month, six months, then as necessary.

Dr Piotr Wozniak, the driving force behind a software package called Supermemo that supports systematic reviews, has carried out the most thorough analysis of the scheduling problem that I have seen and his software has much more sophisticated formulae for determining the optimum intervals. The intervals in Supermemo version 2 are explained on his website and they are 1 day, 6 days, then for subsequent intervals take the previous interval in days and multiply it by a number called the easiness factor (EF). The EF is a number between 1.3 and 2.5 and initially every item starts with an EF of 2.5. As the learner attempts the question he/she enters a number between 0 and 5 representing the quality, q, of his/her answer. The EF is then updated as follows: new EF = max(1.3, old EF + 0.1 − (5 − q) × (0.08 + (5 − q) × 0.02)). If this does not result in a whole number of days then round up. If at any time the quality rating is less than 3 (i.e. you give the wrong answer) the intervals start again as if the item had been learned anew. Piotr found that if the interval did not increase by at least 30% then the reviews were tediously repetitive and this was usually a result of a badly worded question. Subsequent versions of the software have employed even more complicated formulae, though still driven solely by the number of repetitions and the quality of answers.

I recommend deciding when you will next review some material immediately after each review of that material, and flexibly taking into account a number of factors that are usually ignored in fixed schedules. These factors are as follows:

If you are unsure of the intervals to use then Professor Mace's schedule (1 day, 2 days, 4 days, 8 days, etc) is a good starting point. You will soon learn what intervals suit you better in different conditions.

There are a number of ways to keep records of reviews done and scheduled. You can get software to do it, use a spreadsheet, or keep records on record cards in a filing box. You can also move the notes themselves and write dates on them before filing them in date order of next review. Whatever method you choose make sure you do not have to look in several places to find all the scheduled reviews for the day.

Efficient reviews

Rereading is much less effective as a means of refreshing your memory than recalling the memory. The advantage of recalling is so great that it is worth the extra effort. This is another reason for not leaving too much time between reviews.

However, some materials are hard to turn into a memory test without a lot of effort so you may have to compromise.

Many materials can be turned into a very thorough test, either a Question and Answer form or a structured list (i.e. multiple levels like a document outline).

For material in Question and Answer form the ideal way to administer the test is to put the questions in a random order (i.e. a different order each time you do the test) and go through them once. Then go through the ones you got wrong a second time, and again for those you got wrong the second time. This procedure prevents you becoming dependent on irrelevant cues from other items and keeps your effort focused on the most difficult items.

A convenient and portable way to do this is to write the question and answer on opposite sides of cards. Start by shuffling the cards, and when you get a question wrong put it in a separate pile for retesting. If you decide that some items in the pack are so hard or so easy you want to put them on a different review schedule from the others, or stop reviewing them altogether, just take those cards out of the pack.

Alternatively you can use software. Free drilling programs are available on the Internet, and of course there is Supermemo.

A longer list of items can be tested by revealing it one line at a time to check your answer as you go. This is slightly less effort than trying to recall the whole list without a memory crutch, and more realistic. If you were using a list as the outline of an essay for an exam then writing down the items in the list would provide the same kind of memory support, though not the reassurance of knowing you were right.

Again, revealing a list one item at a time can be done with cards or on a computer, where I suggest just using an ordinary word processor and scrolling down one line at a time.

Some professional exams rely on multiple choice questions and to perform well on them there is no better technique than doing a large number of questions from past papers. Keep a list of all the questions and each time you attempt a question record your result on the list. A reasonable target to show mastery is to have answered the question three times in a row without error, with significant intervals (i.e. at least a few days) between each attempt at the question. When you have shown mastery of a question stop attempting it and spend your time on the more difficult ones.

The Supermemo website is here.

Anki is a fairly typically software program for testing yourself. These programs are very simple to write. I wrote one for myself about 20 years ago, but with a very crude user interface. Today it would be much easier.

If you like scientific papers and have access to a suitable psychology library then you may find the collection of references provided by Work-Learning Research quite helpful. It's a good site with much interesting material, including several papers specifically on the "spacing effect" as it is known.

About the author: Matthew Leitch has been studying the applied psychology of learning and memory since about 1979 and holds a BSc in psychology from University College London.