Overlearning is to teaching what cramming is to revision.

Overlearning is continuing practice after the first success, and is very common in maths lessons.

Until 2006 there were no published experiments on the efficacy of overlearning in maths.

The method almost all maths teachers use, almost every lesson to ensure that the learning of the lesson is retained by the learners was not tested, until 2006. A little overlearning is required to ensure the first success isn’t a lucky guess, but after a little more practice, just how effective is overlearning?

In 2006 Rohrer and Taylor published two well designed experiments on the efficacy of overlearning - the results are summarised below.

Overlearning doesn't work well in the long term

The learners were taught how to calculate the number of permutations of letter sequences, with at least one repeated letter

None of the learners could answer the pre assess questions in the pre assess tests.

The learners were split into 4 groups.

Some learners were given 9 practice questions, others were given 3 practice questions.

Some learners were tested after 1 week, others after 4 weeks (not both).

The results are summarised in the table below.


after 1 week

after 4 weeks

9 practice questions



3 practice questions



The extra 6 practice questions that half of the learners did made no appreciable difference to their retention i.e. overlearning didn't significantly improve performance.

Distributed practice works better than overlearning in the medium-long term

In a similar experiment, this time comparing one or two blocks of practice, we see that overlearning is perhaps slightly more effective than distributed practice if the test is in one weeks time, but significantly less effective if the test is in 4 weeks time. 


after 1 week

after 4 weeks

10 practice questions at one time



5 practice questions, wait a week, 5 more practice questions



Rohrer, D., & Taylor, K. (2006). The effects of overlearning and distributed practice on the retention of mathematics knowledge. Applied Cognitive Psychology, 20, 1209-1224.

In the experiments the distributed practice was only separated over 2 sessions, with timely practice we distribute the practice over many more sessions and over a much longer time period. 

Overlearning, retrieval practice and timely practice mp4