Is it Better to Learn Something in Small, Frequent Chunks of Information?

Answer by Robert Frost, author of training course and manual on the learning process, on Quora

It is better to learn small chunks of information, frequently, than big chunks, infrequently.

I will explain by presenting several ideas from experts on learning and then combining them.

In 1956, a cognitive psychologist, from Harvard, named George A. Miller introduced a concept in the journal Psychology Review. That concept has become known as “Miller’s Magic Number” or “Magic Number Seven (Plus or Minus Two)“. In that paper he presented the idea that people can only store seven (plus or minus two) chunks of information in their short term memory. Notice that the word chunks is used, and not pieces or bits. When Miller talked about chunks or “chunking” he talked about our ability to merge a number of discrete pieces of information into a smaller number of chunks. For example, we don’t treat a 10 digit phone number as 10 chunks of information because we combine the three numbers that make up the area code into a single construct and we may chunk four digit numbers as two two-digit numbers.

A related concept called cognitive loading builds upon these restrictions of our minds. Cognitive Load Theory says that the amount of information and interactions that must be processed simultaneously can either under-load or overload the finite amount of working memory. If overloaded, all elements must be processed before meaningful learning can continue. The more a person has to learn in a shorter period of time, the more difficult it is to process that information. Researchers such as Paul Chandler and John Sweller have written extensively on the implications of cognitive load theory on the format of instruction and learning.

Richard E. Mayer is an educational psychologist with more than 390 publications, including 23 books. He has developed a set of learning principles. One of those is the Segmenting Principle. That principle states that:

“People learn better when a complex continuous lesson is broken into separate segments. Examples include breaking a complex figure into two or more smaller figures dealing with different parts of the original one; presenting one graphic at a time rather than putting multiple graphics in the same figure or breaking a continuous presentation into short chunks that can be paced by the learner. The learner’s working memory is less likely to be overloaded with essential processing when the essential material is presented in bite-size chunks rather than as a whole continuous lesson.”

Jerome Bruner is one of the founders of constructivism. His book The Process of Education led to significant experimentation and educational reform during the 1960s. Bruner’s theory of instruction identifies four characteristics of effective instruction (readiness, content structure, sequencing, and reinforcement). Combined, these principles lead to the idea of the spiral curriculum. Spiral learning refers to the idea of revisiting basic ideas over and over, building upon them and elaborating to the level of full understanding and mastery.

Researchers like Lauren Resnick, write that studying what actually happens in the brain – specifically the growth of dendrites and neurotransmitter receptor sites and the formation of neural networks when we think and learn (form easily accessible long term memories) – reveals two fundamental ideas:

1) Learning occurs because of repetition
2) Students must connect new knowledge to previous knowledge in order to learn.

So, putting all of that together, we can deduce that learning will likely be most efficacious when:
1) It occurs in small chunks that can make it through the bottlenecks of short term memory and cognitive load and those chunks are designed to build upon each other.
2) Those series of chunks build upon each other by calling into use the material learned in earlier chunks, providing both repetition and connection opportunities.

This question originally appeared on Quora: Is it better to learn something in small, frequent chunks of information? More questions: