Assignment 1: Models of Memory 1 & 2

Memory is vital to our survival. In psychology, memory covers the following processes:
 >Encoding – whereby information is changed into a form that the system can cope with
 >Storage – for a particular length of time, in a particular form, in a store with a certain capacity
 >Retrieval – getting information out of a memory store


ATKINSON AND SHIFFRIN’S MULTISTORE MODEL OF MEMORY

Atkinson and shiffrin’s theory was that memory was made up of three stores. All three differ in their encoding, storage and retrieval characteristics.

The sensory information store (SIS)
Sperling (1960) thought that information in the SIS was from a feeling. He found this through showing three lines of four mixed numbers and consonants for a brief period of time, then played a note to prompt them to recall what they had seen. The volunteers could do this easily if they immediately recalled what they had seen, however the image faded quickly. In this way, Sperling suggests that the SIS holds between 5 and 9 items.

The short-term memory (STM)
Information that needs to be remembered or used for longer than the SIS is stored in the STM. The going assumption is that information is stored in the form of images, sounds or meanings. Information in the STM is kept ‘alive’ by rehearsal.

Two theories that have led us to this:

Peterson and Peterson (1959) told participants a consonant trigram (three letters) followed by a large number. Then to prevent recollection, the participant had to count backwards in threes from the number and then had to recall the trigram. Participants found it impossible to recall the trigram after 18-30 seconds which indicated that the lifespan of the STM is 18-30 seconds

Miller gave participants a length of consonants and numbers, varying in length, and asked them to recall them in chunks, (e.g. 3-4-3). The amount that all could recall was between 5 and 9 items. This proves that the STM can only hold between 5 and 9 items.

The long-term memory (LTM)
Information, when rehearsed enough times, moves from the STM to the LTM. Information in the LTM is encoded in many forms; knowledge, facts, beliefs, pictures, skills, language, musical knowledge etc. The LTM seems to have an indefinite life span and an unlimited storage capacity.

Further evidence supporting the multi-store model
Murdock (1962) asked participants to learn a list of words. The words presented earlier in the test were put into LTM (primacy effect) and the words shown towards the end of the test were put into STM (recency effect). Words shown in the middle of the test were had been there too long to be put into STM, but not long enough to be recorded by LTM.

Milner (1967) quoted physiological evidence from studies of brain-damaged, amnesic or alcoholic people who have a good LTM but very poor STM. Some stroke victims and elderly people also have STM problems. People suffering from alcohol-related Korsakoff’s syndrome may have problems with LTM but not with STM, although the pattern is not always clear.

Summary of Atkinson and Shiffrin’s model of memory
The model suggests that information is transferred to the LTM by rehearsal but this has been proven to be unnecessary.
We may remember certain things because they are funny, interesting or relevant to us in some way (Eysenck and Keane 1990). Atkinson and Shiffrin also thought that everything was held in the same LTM. This theory, other psychologists have argued over.

Ø Paivio (1971) argued that some of our LTM consists of mental images from all our senses.
Ø Tulving (1972) distinguished episodic memory for personal events and semantic memory for knowledge about the world.
Ø Cohen and Squire (1980) suggest we have declarative memory for storing things we know and procedural memory for storing knowledge about how to do things.

Alternatives to the Atkinson and Shiffrin’s two-process model

Baddeley and Hitch’s working memory model.
Baddeley and Hitch (1974) elaborated on the idea of STM. Their theory was that STM was not only for recording information, but using it at the same time (processing memory), hence the name ‘working memory’. ‘Working memory’ consists of three parts:

The central executive decides how to share out the limited resources of STM. It decides what is and is not to be attended to. It is also ‘modality free’, this means that it deals with both auditory and visual stimuli. The central executive has a limited capacity and deals with tasks that are cognitively demanding.
An articulatory loop deals with verbal information and is essentially a rehearsal system. It is likely that the capacity of this system is limited to that which can be read aloud in approximately two seconds (Baddeley et al 1975).

A visuo-spatial sketchpad can hold and rehearse visual and spatial information. Baddeley and Lieberman (1980) suggested that the visuo-spatial sketchpad relies more on spatial coding that visual coding.

Summary of the working memory model
The important difference between the two-process model and the working memory model appears to be in the role of verbal rehearsal. In line with research findings that have cast doubt over the importance of verbal rehearsal, its role in memory is reduced to the articulatory loop only.
The working memory model can explain how, in brain-damaged patients, selective defects may occur in short-term memory. An updated version of the model has been proposed by Baddeley (1986). This sees the articulatory loop as compromising a passive phonological store (concerned with speech perception) and an articulatory control process (concerned with speech production). The revised model was better able to explain some of the neurophysiological evidence that did not fit the original model.

CRAIK AND LOCKHEART’S LEVELS OF PROCESSING THEORY
Craik and Lockheart (1972) disagreed with Atkinson and Shiffrin’s idea that memory consisted of separate stores. Instead, they suggested that memory depends on how we process the information when it comes in.
Shallow processing takes two forms. Structural (or ionic) processing is when we encode only the physical qualities (appearance) of something. Phonemic (or acoustic) processing is when we encode its sound. Shallow processing only involves maintenance rehearsal and leads to fairly short-term retention of information.
Deep processing involves semantic processing which happens when we encode the meaning of a word and relate it to similar words with similar meaning. Such elaborative rehearsal leads to longer-term retention.

Summary of the levels of processing approach
The levels model seems to describe what is happening rather than explaining why deeper processing leads to longer lasting memories (Eysenck and Keane 1990), e.g. deep processing may lead to enduring memories simply because it takes more time and effort than shallow processing. It is also difficult to test the model by controlling what kind of processing people use. Various modifications have been suggested:

Elaboration - Craik and Tulving (1975) found that it was not just depth of processing that affected retention but also the degree of elaboration a person carried out, e.g. elaborating ‘table’ by thinking of tables of different sizes, made of different materials, for different uses makes it more likely that we will remember ‘table’.

Distinctiveness – Eysenck (1979) suggested that if a memory trace is made distinct from other similar ones, it will not get confused with them. We remember some events, such as personal successes or disasters, because they stand out as being unusual or distinctive.

Context – Tulving (1979) suggested that the setting (context) in which something is learned is encoded along with the material to be remembered. If we learn something in a particular context than in a different one reguardless of how deeply we processed it.

Personal relevance – Rogers et al (1977) found that participants who processed words in terms of whether they applied to them in some way (e.g. Do you own one of these? PARROT) remembered them even more than semantically processed words.

End.

iMMense.