Psya 1 - Cognitive Psychology - Unit One the Multi Store Model of Memory

PSYA 1 - COGNITIVE PSYCHOLOGY - UNIT ONE – THE MULTI STORE MODEL OF MEMORY

LEARNING OBJECTIVES - YOU WILL BE ABLE TO:

1. Describe the multi-store model of memory.

2. Understand the concepts of capacity, duration and encoding.

3. Explain the strengths and weaknesses of the multi-store model.

Psychologists are interested in factors affecting memory. Various studies have helped to support theories about how memory works. These theories can be shown in the form of ‘models’ or flow charts about how memory is thought to work. The following is one of the first and most influential models of memory.

1. The muti-store model of memory – Atkinson and Shiffrin (1968) Different ‘stores’ hold information. The model describes three separate stores through which information passes. At each of these stages (especially the first two), information can be lost (forgotten). PSYA 1 - COGNITIVE PSYCHOLOGY - UNIT ONE – THE MULTI STORE MODEL OF MEMORY The model proposes that information enters the system from the environment and first registers in the sensory memory store where it stays for a very brief period of time before either being lost (decay) or being passed on the short-term store. Attention must be paid to the information for this to happen. The short-term store has a very small capacity which means it can only hold very small amounts of information at any one time. The memory traces held here are quite fragile. The duration is also very short. That is, if information is not rehearsed and used it will be lost within a few seconds.

Items in the short-term memory are usually held in the form of sounds although other kinds of encoding are possible.

If the material is sufficiently rehearsed, it is passed on to the long-term memory where it can stay for a life time. It can be lost through damage to the brain and/or through the processes of decay or interference.

This is called a structural model because it focuses on the storage components of the memory system. However, Atkinson and Shiffrin also described some of the control processes required to manipulate and transform the information as it flows through the system. These include encoding, retrieval and rehearsal. One of the most important processes is rehearsal whereby information can be circulated within the STM and passed on to the LTM.

Terminology: Encoding - the form of representation used e.g. visual, acoustic, semantic Duration - how long an item lasts in the store Capacity - the total amount of information that can be held at any one time

These properties can be studied experimentally to find out whether the proposed stores have different characteristics.

The evidence summary – MAKE SURE YOU KNOW WHAT THE STUDIES WERE INVESTIGATING, I.E. THE AREA OF RESEARCH AREA OF RESEARCHERS METHOD OF RESEARCH RESEARCH SHOR Sensory store Sperling (1960) T- Duration Peterson & Peterson (1959) TERM Capacity Jacobs (1887) Laboratory MEM Miller (1956) experiments ORY Encoding Conrad (1964) LTM Duration Bahrick et al. (1975)

Encoding Baddeley (1966) BE AWARE THAT MOST OF THE FOLLOWING ARE LABORATORY EXPERIMENTS. THIS MEANS THAT INDEPENDENT VARIABLES ARE MANIPULATED AND DIFFERENCES BETWEEN CONDITIONS ARE COMPARED. CONTROL IS HIGH. PSYA 1 - COGNITIVE PSYCHOLOGY - UNIT ONE – THE MULTI STORE MODEL OF MEMORY 1. Sperling (1960) – sensory memory Procedure Showed participants 12 letters in 3 rows (4 in each row) for 50 milliseconds. Findings People remember seeing more than they can recall, but the image fades during the time it takes to report back four of the items. It’s a bit like trying to read the credits that roll up on the screen at the end of a film. While you are paying attention to one name, others are disappearing off the screen. Conclusion Decay happens very quickly in sensory memory (ie. Duration is short) and capacity is very limited.

2. Peterson and Peterson (1959) – duration in short-term memory Procedure Participants were: -shown a consonant trigram (no vowels so that it can’t be transformed into a sound eg. CXK) -asked to count backwards out loud in threes from a specified number (e.g. 451) to stop them rehearsing the trigram -after intervals of 3,6,9,12,15,18 seconds were asked to stop counting and to repeat the trigram. This was repeated using different trigrams on each presentation. Findings Participants were able to recall about 80% of trigrams after a 3-second interval without rehearsal but their recall became progressively worse as the time intervals lengthened until, after 18 seconds, they could recall fewer than 10% correctly. Conclusion Information decays rapidly from STM when rehearsal is prevented. Duration is limited.

MAKE SURE YOU KNOW WHAT THE DIGIT SPAN IS 3. Jacobs (1887) – Digit Span technique -Capacity in STM - Procedure Presented digit strings to participants which increased in length each time. Findings Participants could recall about seven digits on average. This was supported by many subsequent studies such as Miller.

4. Miller (1956) ‘The Magical Number Seven’ Capacity in STM Miller proposed that we can hold about seven items in our STM, but that there is a range of capacity between five and nine items. These ‘items’ can be chunks of information rather than individual letters or numerals. For example, the following list of letters can be chunked: X G U W Z S P J Q L T B F M K becomes XGU WZS PJQ LTB FMK If encoded as chunks of information, more digits overall were recalled. Even though the chunks are not meaningful they impose a rhythm which seems to make it easier to recall them.

If the chunks are meaningful, capacity may be improved even more, e.g. B I T K E G S U N L A W T O Y becomes BIT KEG SUN LAW TOY Conclusion Chunking information increases capacity in STM PSYA 1 - COGNITIVE PSYCHOLOGY - UNIT ONE – THE MULTI STORE MODEL OF MEMORY 5. Conrad (1964) – Encoding in STM

Procedure Participants were shown a random sequence of six consonants projected on a screen in very rapid succession. There were two conditions: 1. letters were acoustically similar (e.g. B, G, C, T, D, V) 2. letters were acoustically dissimilar (e.g. F, J, X, M, S, R) Immediately after the presentation, participants were asked to write the letters down in correct serial order.

Findings Participants found it more difficult to recall strings of letters that sounded the same than letters that sounded different. They often substituted one for the other, e.g. V for D etc.). Conclusion Visually presented information was encoded acoustically and caused confusion where it was similar.

6. Bahrick et. al. (1975) – Duration in LTM

Procedure Tested the memory of 392 graduates of an American high school for their former class mates using pictures, matching names to pictures or using names with no picture cue. Findings Participants recalled very well up to about 34 years. Recall was better on recognition tasks rather than on free recall (Ii.e. when pictures or names were given). After 47 years there was less recall on all tasks. Conclusion LTM has a very long duration. Accurate recall is better with recognition tests rather than free recall. Depth of learning affects duration of memory.

7. Baddeley (1966) - Encoding in LTM

Procedure Participants were presented with a random sequence of ten words There were four lists:  Acoustically similar (e.g. mad, map, mat, cad, cap, cat)  Acoustically dissimilar (e.g. pen, cow, pit, sup, day)  Semantically similar (e.g. tall, high, broad, wide, big)  Semantically dissimilar (e.g. foul, thin, late, safe, strong). Each list was presented four times and then recall was tested after a 20-minute interval. Rehearsal was prevented.

Findings Acoustic similarity had no effect on recall but words that similar in meaning were poorly recalled. Conclusion LTM codes were mainly semantic (for meaning). PSYA 1 - COGNITIVE PSYCHOLOGY - UNIT ONE – THE MULTI STORE MODEL OF MEMORY 8. Glanzer and Cunitz (1966) – Primacy and Recency effect – the distinction between STM and LTM (functional dissociation) Procedure Participants were given lists of words presented one at a time. They were then asked to freely recall. There were two conditions: 1 participants were asked to recall immediately 2 participants were given a distractor task after the presentation of the words and had to count backwards in threes for 30 seconds before recalling Findings Condition 1: the first and last items were recalled better. Condition 2: words from the last part of the list were not well recalled.] Conclusion The interference task (counting backwards in threes) had displaced the last few words in the list from the fragile STM but the first words had already been rehearsed and passed into the robust LTM.

DETAILS OF WHY THE FOLLOWING BRAIN DAMAGE OCCURRED WILL NOT BE REQUIRED. IT IS ENOUGH TO BE ABLE TO REFER TO HOW THE DAMAGE AFFECTED ONE ASPECT OF MEMORY WITHOUT AFFECTING THE OTHER. BE AWARE THAT THESE ARE CASE STUDIES NOT EXPERIMENTS. 9. Milner (1966) – the case study of HM HM suffered from epilepsy and had parts of his temporal lobes and hippocampus removed. His epilepsy was alleviated, but he suffered severe memory deficits. His IQ remained above average. He was able to recall events in his early life but was unable to remember events for about ten years before the surgery. He could not learn or retain new information. He could remember approximately six numbers in the order they had been presented. However, he repeatedly read the same magazine without realising he had read it before. He was unable to recognise the psychologists who spent long periods of time with him. Conclusion HM had a normal STM but his LTM was defective. This suggests that they are independent stores.

10. Shallice and Warrington (1970) KF KF sustained brain injuries after a motorcycle accident. He appeared to have an intact LTM. He could learn new information and recall stored information. However, he had a digit span of one. Conclusion KF’s capacity in STM was damaged, but his LTM functioned well.

IT IS ENOUGH TO KNOW THAT THERE IS SCIENTIFIC EVIDENCE E.G. BRAIN SCAN RESULTS, WHICH SHOWS THAT DIFFERENT AREAS OF THE BRAIN ARE INVOLVED IN DIFFERENT ASPECTS OF MEMORY SUCH AS STM AND LTM 11. Squire et al. (1992) – brain scanning techniques Procedure Positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) was used while different memory tasks were given. Findings The hippocampus is active in LTM tasks whereas areas in the pre-frontal cortex are activated for STM tasks. Conclusion STM and LTM are distinct and operate independently. PSYA 1 - COGNITIVE PSYCHOLOGY - UNIT ONE – THE MULTI STORE MODEL OF MEMORY

Evaluation of the Multi-store model of memory

MAKE SURE YOU CAN CONFIDENTLY DISCUSS THE HIGHLIGHTED ONES

 The multi-store model has made an important contribution to memory research. The information-processing approach has enabled psychologists to construct testable models of memory and provided the foundation for later important work.

 There is plenty of evidence to support the distinction of a distinction between short-term, temporary, limited-capacity store and a more robust and permanent long-term memory. Case studies of brain damaged patients provide evidence for a separate stores which function independently (e.g. KF’s STM was impaired while his LTM was intact).

 The model is over-simplified and fails to reflect the complexity of human memory, for example it places emphasis on the amount of information that can be handled at any one time, but takes no account of the nature of things we have to remember. Some things are easier to remember than others because they are more interesting, more relevant, funny etc.and do not need to be rehearsed.

 There is considerable evidence that simple repetition (as suggested by the multi- store model, is one of the least effective ways of passing on information. Craik and Lockhart found that things are remembered better if they are processed semantically (in terms of their meaning). Some things are remembered because of their emotional impact and do not need to be rehearsed. Neither can the MSM account for how some people develop strategies for remembering things.

 Some evidence points to an interaction between STM and LTM rather than a linear model. For example, KF (see case study) had a severely impaired STM and yet his LTM appeared to work quite normally. This suggests that the flow of information through the memory system is interactive rather than sequential.

 Much of the supporting evidence for the multi-store model comes from artificial, laboratory studies, which might not reflect how memory works in everyday life. It is sometimes possible to interpret the results of such studies in different ways. It is also the case that different experimental techniques can yield different results. For example, when acoustic coding is prevented by asking participants to repeat a meaningless chant (‘la la la’), Brandimonte et al. (1992) found that visual coding can be used and can even be more effective than acoustic coding.