LTM Storage and Retrieval

Effective LTM Storage

It takes time for consolidation to happen; for the brain to modify itself and concrete a memory (syntactic changes in the hippocampus need to take place). Studying the night before is not useful for LTM storage, only for STM.

Number of Revision Sessions

Learning/Rehearsing a concept x times on day 1 is not enough to recall it on day 2, but more repetitions does decrease the time required to relearn it the next day.


Time Between Revision Sessions

The Spacing effect is the phenomenon whereby the ability to recall information is improved when they are revised a few times over a spaced out period of time.

Bahrick's 1993 study of Spanish showed the same phenomenon with foreign language acquisition:


Depth of Processing

There are different levels of mental processing required to comprehend different types of information/requests. The link between this level and the ease of recall is known as the levels of processing effect (identified by Craik and Lockhart in 1972) - essentially a deeper level of processing leads to easier retention.




Ebbinghaus hypothesised that the speed at which forgetting happens is dependent on a number of factors, including how meaningful the content is, the representation and various physiological states (e.g. stress, sleep).

Meaningfulness can mean things like "PED" become 'pedal' or "BOL" becoming "ball" - anything to give arbitrary things a meaning we can relate to.

Relating Information to Existing Knowledge

Information relating to existing knowledge is easier to comprehend; it allows us to build on our mental picture and fit new content in to that rather than constructing a new starting point.


Organising words into different categories provides us a category framework to build off - relating concepts and trigging memories of multiple because of linkages. Mandler's 1967 study showed the link between organising into categories and improved recall.

Semantic Network Models are a description of the way we create these categories - depicting semantically similar words connected in a network. This technique can lead to false remembering, by believing to have heard a closely related word.


Elaborating on a concept or image by creating a richer scene with deeper processing allows it to stick deeper and creates more trigger possibilities by linking concepts together.

"Palmere et al. (1983) gave participants descriptive paragraphs of a fictitious African nation. There were some short paragraphs and some with extra sentences elaborating the main idea. Recall was higher for the ideas in the elaborated paragraphs."

Mnemonics and Rhyme

Mnemonics and rhyme can be useful techniques for remembering lists, but are mainly useful in that they cause us to elaborate and create a rich visual image.

Relate to Personal Experience

Symons and Johnson (1997) described the self-reference effect as the level on which we personally relate to information affecting the way we store/encode information.

Effective Retrieval of Information

Context Specificity (State Dependent Learning)

Extrinsic Context (Physical Environment)

By manipulating the context of retrieval and encoding we can study the effect of context on memory. Godden and Baddeley (1975) put subjects either on land outdoors, or under water (where the environment is very noticeable and salient) and gave them a word list to learn, and then tested their recall in each environment.

The results showed that the relationship between same learning and same testing context was good, and swapped around poor.


Smells are also very important to information retrieval (a part of the environment) - it's interesting to try introducing a distinctive smell for each subject and then wearing that scent in your exam.

Other Physiological Contexts

Mood and mental state has a distinctive effect on our ability to recall and learn things. The same thing happens with substance use.

In Conclusion

  • Learning and testing contexts match = better recall
  • Forgetting is often due to context change
  • This is called state dependent learning (our physiological state affects us - mood, environment, drugs/etc).

Two ways to Test Memory

  1. Recall tests - free response test - answers are not restricted by anything in the question, e.g. write down all the words you saw in the study list
  2. Recognition - forced choice test, which of these words have you seen before in this list, etc.

Both recall and recognition tests are declarative (we can describe them), and recognition is easier than recall (wins out) as memory is better when the test conditions match the encoding ones (e.g. looking at a word).

Retrieval Cues

  • Recognition: the word in the recognition test is a retrieval cue, and it becomes a simple pattern matching task
  • Recall: there is no retrieval cue of specific memories; you have to use the question as a cue rather than a word and it's hence much harder than a yes/no match.


The problem with recognition tests is that you end up with leading questions; a big problem in witness declarations.

Implicit memory/Priming

Priming involves preparing people to expect a particular thing by previously activating that concept in the brain with images or words or cues.

Priming is implicit - there's no deliberate remembering involved, no conscious effort. It's procedural, and we needn't know that we've been primed for it to work.


Priming us with images of rabbits flashed throughout the slides means that ambiguous duck/rabbit images will be interpreted as a rabbit, as it activates the memory of a rabbit.


Priming with Rhyming

We can prime people with rhyming responses


Testing Priming

We can prime a group and then compare to a control group over related puzzles.

  • Solve britab - if you've been primed with rabbit you'll do it faster
  • Is rabbit a word - 50ms response for yes (faster than non-primed)
  • Complete this word r_b___ = rabbit