The Psychology of Quality and More
CHAPTER 2 : Psychological Factors
2.7 Short Term, Working and Long Term memory
Your memory has three significant components, which equate quite well to a computer model.
Short term memory is equivalent to cache memory. It is fast but volatile, remembering things for only a few seconds. It is severely limited in size, holding only about seven items. All items entering memory start here. If you are distracted, even for a moment, you can forget what is in short-term memory. How often have you been introduced to someone, then immediately forgotten their name, as you get distracted by other verbal or visual detail?
Working memory is equivalent to main computer memory. Its capacity and retention are better than short term and less than long term memory. It is where information from short and long term memories is processed during problem solving.
Long term memory is like disc storage. It is slow moving, but is very retentive. It has almost infinite capacity.
We read and understand things by first moving them from short-term memory into working memory, where is then a pause whilst they are examined, pattern-matched and differentiated from existing information, then integrated into long term memory. The more processing we do to improve this integration, the more likely we are to remember to item later.
When an item is in short-term memory, it fades very quickly, and has to be constantly refreshed via maintenance rehearsal in order to stay there. If the item is familiar, then it can be recalled from long term memory. Otherwise it may have to be replaced, which means that item must be visible. In a program reading context, this means that it is very helpful if all current information is visible, typically being on the same page.
When we have a list of things to remember, the early items start with a 'clean slate' and get into long term memory reasonably easily. The final items also tend to get remembered, as there is time to integrate them in the vacuum after the end. The items in the middle are less likely to be remembered, unless the list is read slowly enough to integrate all details. Coding-wise, this implies that important details are better at the beginning or the end of a list.
And the big